I think people are being a little too hard on the author. Startups in healthcare are hard. Been there, done that, have the t-shirt (literally, that's all that left). I even made the same fatal mistake - not identifying who the buyer really is. And when I say identifying the buyer, I don't mean it in that hand-wavey, vague way like "doctors" or "insurance companies". The starting point has to be something like "ophthalmologists in small practices (1-10 doctors) in New England who are trying to acquire new patients through social media."
And here's something that's absolutely critical for all engineers trying to build a startup in healthcare to really understand. Healthcare is so alluring because tech people are so idealistic. We think, "Wow, we we can write some code, and then save lives!" And that's sometimes true. But when you are thinking of your customer, revenue model, sales strategy - all that business stuff - the mistake we make is thinking healthcare is different and the same rules don't apply. We think if we can save lives, we can make a business. But the key insight is that you have to take a step back and just treat healthcare like any other for-profit business. Doctors and hospitals care about what generates revenue. Insurance companies want to save money. Pharma companies are looking to advertise to new customers. It's no different than other industries. The psychiatrist's reaction to the sales pitch is the classic thing a doctor will tell when you're not helping their bottom line. As I was reading the post, I could almost predict how she was going to respond.
And if you figure that business stuff out, who knows, you might just build a healthcare business that saves a life or two.
From a business standpoint, saving lives can actually be expensive. The longer someone lives, the more healthcare they consume.
Health insurance companies are very aware of this, which is why they so frequently deny coverage for various life-saving cancer treatments. If you die while waiting for treatment, not only do they save the cost of the treatment, but also the entire cost of ongoing screening / care during remission.
The longer you live the more monthly premiums you pay, so ideally they do want you to live, but they don't want you to use any healthcare services (or use less services) while you live.
No, the moment you are diagnosed with a chronic or life threatening disease, your premiums no longer cover the medical cost. Otherwise what is the point of insurance. Moreover, many regulations prevent them from raising future premiums. So kicking you off the insurance plan or letting you die is indeed their primary concern
This is incorrect. How can they kick you off the insurance plan? Also, the ACA has provisions to prevent one insurer from having to shoulder the costliest patients via some form of cost sharing with all the others.
They can investigate whether you failed to report any preexisting conditions, or otherwise violated terms of your insurance. Things that would seem to you to be a technicality may be to their lawyers a breach of contract. Do you have a lawyer who is up for that fight?
ACA, since 2010, made it so health insurance companies cannot discriminate for pre existing conditions.
If you notice, when you search for insurance quotes on healthcare.gov, it only asks you for age, location, and smoking status. And gender, but I don’t think gender can be used to price the insurance any differently.
It boggles my mind that people keep bringing up pre existing conditions a decade after they have ceased to be a thing. Also, your link doesn’t work.
A better way to put it is that health insurers don't want you to get "sick". So anything to prevent that is usually money well spent. That's why vaccines are so easy to get.
"This is precisely why we need universal healthcare. The incentive structure of private markets don't properly value human life."
Government-run healthcare isn't a panacea that will 'value human life' any more than the insurance companies. You are still a number in a database somewhere.
Most new drugs come out of the US and socialized governments create generics at a fraction of the cost, because they don't have to put any money into R&D.
The US is the back-bone of the medical system for the rest of the world and also has the best medical schools and best quality hospitals (the majority of people with money come to the US for life-saving surgeries..this has to tell you something).
We need to push the true costs of things to the patient, which will force the price down...not replace the existing inefficient layer of the insurance companies with an even more inefficient layer called the government.
This has worked with many cosmetic surgeries, which used to be very expensive, but aren't covered by insurance.
Insurance should only cover rare surgeries that can't use the free market to reduce the costs.
I can't think of anything that the government runs better than the private sector. Healthcare is no different.
> The US is the back-bone of the medical system for the rest of the world and also has the best medical schools and best quality hospitals (the majority of people with money come to the US for life-saving surgeries..this has to tell you something).
This just means American consumers end up subsidizing health care R&D costs for the rest of the globe — purely because our government prevents us from collectively bargaining the way pretty much every other country does.
The people coming to the US for the “best quality” are generally going to doctors / surgeons in private practice, most of whom do not take insurance and can see you immediately. There is an almost entirely separate (and much higher quality) medical system available to those who can afford it. The system the rest of us belong to ranks right behind Morocco and Costa Rica in terms of outcomes (not joking — this is according to the WHO). We come in dead last among “first world” countries.
If American consumers stop subsidizing health care R&D costs, what will happen to the R&D? Will it decrease? Will health care costs in other countries increase as they start to bear some of it?
I assume it will continue because the drug companies need products to sell. It's not like they're just going to close up shop, but it may be the end of windfall profits.
The other problem right now is that it's not profitable to create drugs that actually cure anything. It's far more profitable to research drugs that help control chronic conditions (like diabetes, hypertension or Alzheimers) rather than drugs that can cure a condition outright. This is one part of why nobody researches things like antibiotics anymore.
As stated in a thread above, while it is not directly profitable for drug companies to cure diseases, it is profitable for their insurance agency cohorts. The current system does not totally dis-incentivize cures.
The problem is that those drugs never even make it into the pipeline because they don't get investment in the first place. The big drug companies aren't the ones actually developing new drugs; that's done by biotech startups. If a drug shows promise, big pharma buys the company and pays for stage 2+ clinical trials.
A friend of mine works in pharmaceutical marketing, and the big revelation I got from him was that the marketing departments run the drug companies from the top to the bottom. If marketing doesn't see a billion dollar market for a new drug, it won't even be considered for clinical trials. The VC in biotech knows this, so they won't fund startups that don't have this potential.
There was a HN thread on this recently in relation to antibiotics. [1] Antibiotics are a special case for their own reasons, but ultimately if the potential for insane windfalls isn't there, big pharma is not interested.
Between the MRC and NIHR, the UK's NHS puts at least a couple of billions (Sterling) into clinical research each year, making them the largest clinical research funder in Europe. They're not researching generics. Oh and they were one of the original funders of PubMed Europe.
I can think of lots of things the government does better than the private sector, although admittedly outside the US.
> Objectives. We explored whether the United States, which does not regulate pharmaceutical prices, is responsible for the development of a disproportionate share of the new molecular entities (NMEs; a drug that does not contain an active moiety previously approved by the Food and Drug Administration) produced worldwide.
[...]
> Conclusions. Higher prescription drug spending in the United States does not disproportionately privilege domestic innovation, and many countries with drug price regulation were significant contributors to pharmaceutical innovation.
And
> In the United Kingdom, the pharmaceutical industry invests more of its revenues from domestic sales in research and development than do companies in the United States.
The US pays a lot for medication, but a lot of that money is used for things that are not innovation (eg advertising). Countries that have better free markets for medications don't see less spending on innovation, they see more spending on innovation.
Implementing something like what you describe, IMHO, would have the greatest positive impact on the health and welfare of Americans, of anything this century.
I have a friend who is very high up in one of the largest health insurance providers in the US, and I have often debated over the years with this person that same point, that the insurance layer is what creates the model for inefficiency, and abuse at the provider level ($5 Tylenol pill, $10 Band-Aid, minor treatments and return visits of dubious value, etc.), and that if we treated medical care as an ordinary consumer good, outside of the extreme emergencies that are bankruptigly expensive, things would be a lot better. The person I know at the insurance company doesn’t necessarily disagree either.
Take a look at some of the "universal catastrophic coverage" proposals. The ideas is to BOTH socialize and increase market competition for healthcare at the same time. Universal care if you get cancer or for other chronic/terminal conditions. Then have a real market and competition for everything else. Regulate safety but not pricing or delivery. (Many of these plans also call for universal preventative services that drive down costs for the whole system). The Dutch have something closer to this in practice with "universal long-term care insurance" separated from the basic/primary insurance, but I've heard mixed reviews of how this works in practice.
It’s an interesting concept. The primary concern is that most of the costs for catastrophic medical problems will be incurred due to things like obesity, smoking, heavy drinking, dangerous lifestyles, etc. This creates one of two problems (or both?): either I’m left with the bill for others’ bad choices, or individual agency is under state / democratic control (e.g., salty foods banned, dessert banned, drinking banned, etc.).
Neither of those scenarios are ideal.
If the only question was genetic differences, that’s a risk I think a lot of people would be willing to take (e.g., that I might be healthy until 100 and then die in your sleep, theretofore paying constantly for others who have had medical problems for 70 years straight). I myself would be willing to take that risk / pay for that style of universal coverage, if not for the aforementioned choice part (smoking, etc.) playing so heavily into the cost, in practice.
The costs vs. outcomes analysis across health care systems suggest that your personal lifetime expected return on investment (likelihood * (value/cost)) of a universal health care system with a single large national risk pool will be higher than your lifetime ROI in the current US system.
You are probably underestimating your own unknown non-lifestyle associated risk, along with the risks your offspring represent.
It also costs a lot to detect and judge risk factors, manage individual risk pools, assure compliance, etc. You save a lot of money by simply not bothering to risk-qualify eligibility and this offsets the costs associated with carrying lifestyle risks in a single large pool.
This is a shockingly uncompassionate perspective - is it satire?
Assuming it's genuine, it's also misguided. Almost no-one wants to be obese, alcoholic, etc - accessible healthcare will reduce the incidence of all of these, which in turn will reduce the externalities associated with them.
You also don't need further state control to discourage smoking, drinking, junk food, etc - taxation is the mechanism used globally to discourage undesirable behaviour.
Anarcho-capitalism -- er, "libertarianism" -- at its finest. It's a deeply nihilistic worldview.
Almost all these things are biologically addictive, and furthermore the companies selling them have long known about their addictive qualities and then built marketing campaigns to ensure the public was not aware of the risks.
Taxation is how you recognize the externalities of this behavior while maintaining a free-market system.
We’ve been taxing cigarettes at like 100% or something ridiculous for many years, yet just as many people die from tobacco-related cancers now (if not more) as did when cigarettes were much worse for you (no filters, more chemicals, etc.). Those taxes also did not help to alleviate any of the problems in the medical industry; or if they did, they were obviously offset by something else massive, since cancer healthcare expenditures have risen drastically since their inception.
Also, it’s easy to look at something like cigarettes which pretty much everybody would agree are a terrible idea, and say, “fine, tax em”, but what about when it’s sugar, then butter, then red meat, etc. Let people make their own choices instead, and just don’t put the onus on the rest of us to fix the problems of those choices. Banning smoking in public is acceptable because it’s no longer a choice that only affects yourself; it affects anyone near you in a really terrible way. But taxing cigarettes (or anything else) is no solution.
“Consumer driven healthcare” as the academic researchers call it. It is bewildering why we don’t have diagnostic labs doing email marketing campaigns, for instance. This is driven, I think, by an outmoded concept of the security of doctor’s knowledge. Now, doctors are so behind the ball and play defensive to protect themselves from malpractice. The idea that an individual cannot make educated decisions which are market driven (ie price driven) for most mild conditions is incomprehensible.
With respect to health care, you may be able to make such educated decisions, almost no one else can. There isn't enough signal, and almost no one has the ability to distinguish signal from noise, and correctly interpret it.
What happens is that when given the opportunity, business will pollute the market with proxy signals (like brand trust, or fear based selling) that people instinctively respond to.
These proxy signals will be subject to manipulation by well-capitalized actors and will not in general reflect outcomes or value. They will interfere with the fundamental signal, making it even harder to detect and interpret.
Therefore the market will not work, in the sense of delivering the best outcomes for costs.
(Note that I am strictly speaking of healthcare, where even experts are often unsure of causal relationships and the relative value of various therapies. In general I think markets are a wonderful thing. Here is an experiment to try... take a mild condition treatable by OTC medication, such as the symptoms of colds, allergies, etc, and see how long it takes you to identify, characterize and factor the available therapies, and then map this to available products. Evaluate interactions, cautions, effectiveness, risk, convenience, costs, etc. Now ask yourself, where did you get your information? Who did you have to trust? Why did you trust them? Is the basis for your trust and understanding available to the average person?
I've done the experiment. My conclusion is that healthcare large and small is not like other goods.)
You assume medical bureaucracies are needed to protect from self-harm; ironically the current healthcare state is harmful!
I have firm belief in the fact that the market can more efficiently aggregate medical information but above all else create more rapid responsiveness to care. Bad products and actors are always seeded out over time in a society.
I really wish that all markets worked like that. When they work, markets are an amazing mechanism for pricing, driving value creation, allocating capital, and searching the economic and technological solution space.
But they don't work like this all the time, for all goods.
Tetraethyl lead, which quite probably caused material long-term harm to entire generations, is gone because the bureaucrats banned it. The free market would never have phased it out organically in a reasonable time-frame, we know this because it didn't. The market participants fought for its continued use and obfuscated the issues to their advantage.
GlacierMD was explicitly an attempt to filter, aggregate and make available high-quality health-care information (i.e. improve the quality of the signal, which is necessary for efficient market functioning), and it did not find a value proposition in the market.
> GlacierMD was explicitly an attempt to filter, aggregate and make available high-quality health-care information (i.e. improve the quality of the signal, which is necessary for efficient market functioning), and it did not find a value proposition in the market.
Except it doesn't -- what GlacierMD was trying to do is something that's currently done by humans at scale. Which he would have known if he understood the market better. This is a centralized function that is performed much higher up the chain and ends up in clinician / patient hands in the form of clinical guidelines.
This is maybe a useful tool for researchers to be used as one input in the development of clinical guidelines. But the functionality it delivers can be done manually (and will continue to be for liability reasons). In this case, the market worked. The author should be thankful he was able to get out of it only $40k in the hole.
I’m not trying to be divisive, but allow me to counter that all markets need consumer advocacy. High fructose corn syrup has been substituted in certain soft drinks now because of the diligence of consumer activism. I take it as a mathematical certainty that bureaucrats take longer to respond than the fluidity of market economics. I just find it appalling I can’t order my methotrexate medication without having to physically see a rheumatologist, with the man-hours to get that script embarrassing for a digital society.
The government runs a lot of things better than the private sector. Even in the US, the government loosely or strongly manages Police, National Defense, Garbage Disposal, Water managenemt, Power Management, Internet infrastructure, Education, Telephone, Mail and many more. Any service that every citizen needs to live a healthy life is better and more cost effectively run by the government.
Telephone and internet are usually run by private companies no? I think the post meant that the few government run alternatives that exist tend to be better (cheaper, faster, better service etc).
I didn't claim it was. Only that our present standards incentivize fee-for-service and not outcomes-driven, as exemplified by the author's inability to sell his outcome optimizing software. ACA tried to remedy that with preventative care standards, but the current administration is actively dismantling it.
Your model works for cosmetic surgeries because they are cosmetic - optional and not life threatening.
When calculating price, the cost to produce a drug is the last thing on pharmaceutical companies' minds. The question they ask is how much are you willing to pay for something that allows you to continue living. The factors they actually consider: the patient's perceived cost-benefit of continued suffering, the credit/insurance-worthiness of the treatment cohort, and whether the company faces competitive pressures from others selling the drug on the market.
Nothing better depicts the situation than what goes on when pharmaceutical companies merge or acquire the rights to rarely used drugs[1].
The only way to structurally combat this is with the monopsony power of a single payer system.
>I can't think of anything that the government runs better than the private sector. Healthcare is no different.
Government run care has the highest satisfaction rate of any healthcare coverage[2].
> I can't think of anything that the government runs better than the private sector.
Then you are not very good at thinking. There is no theoretical, practical or empirical evidence that this should be true. If you _feel_ this should be true, you should really investigate where that feeling comes from.
Basically your post is just ideology without argument or evidence. All evidence shows that free markets in health care are terribly inefficient. If that doesn't align with your ideological priors, well, adjust your reality.
I am on the fence between Single Payer and your proposed solution of "push the cost to patient to force the price down". I do think the same but the appeal of universal healthcare is similar to the appeal of things like fire department. We don't really always need it but when we do, we don't want to worry about it.
Honestly, as long we get rid of the mafia middleman a.k.a insurance company except for real catastrophic situations, I do agree that costs will plummet. Price transparency is another key aspect.
The problem with the fire department is the freerider problem - if I pay for the fire department’s “protection” but my neighbor doesn’t, and his/her house is on fire, the fire department has to save my neighbor’s house in order to save mine, so the neighbor is not incentivised to pay.
Healthcare doesn’t have this issue (except in emergency services), so public healthcare also doesn’t solve that. It solves problems such as the principal-agent problem, asymmetric information, improves collective bargaining, ...
Hard to say. These are all internationalized firms with a variety of global operations centers. I know GSK does a lot of their research in North Carolina. But that is beside the key argument, which is that the US pays more (in aggregated cost and volume) to these firms, such that the total revenue for a product line during the patent lifetime justifies the cost to both bring the drug to market and pay for a variety of failures that didn't make it to market. I'm not sure if this is true, but it does seem plausible. And even if it is true, it still might be better to, say, fund more university based research instead, or have lower per-unit costs and pay innovation bounties or any number of other schemes.
I have worked in government (twice) and large and small companies, now working in the private sector for the last 20 years. There are structural inefficiencies and perverse incentives in both domains, though they are very different in style and visibility.
>Most new drugs come out of the US and socialized governments create generics at a fraction of the cost, because they don't have to put any money into R&D.
This is a completely false right-wing talking point. It takes 10 minutes to uncover how just totally bullshit this is.
Don't ever repeat again or anything remotely like it (aka the rest of your post).
The government never sees a return in any form on the money taxpayers put forth. But private companies do. Just waiting like vultures to snag whatever makes it through the proof of concept phase and claim credit.
>I can't think of anything that the government runs better than the private sector.
ISP's? Roads? Social services? Schools? Emergency Services?
Your second link has a pie chart that claims that pharma companies spend around 200x as much on advertising as clinical trials. Do you believe that is true?
This has been known for a very long time in pharma circles. Marketing budgets are several orders of magnitude higher than R&D budgets. An extra dollar spent on marketing has much higher returns than an extra dollar spent on R&D. Most of the real "hard science" R&D is paid for with NIH grants.
There are definitely lots of problems with the pharmaceutical industry, several of which are mentioned in that BBC article. But making hyperbolic claims that they essentially don't do any R&D isn't helpful.
>pharmaceutical company spending on marketing far exceeds that of its research budget
Imagine living in a society where supposedly rational, intelligent people hold opinions that don't withstand the test of a few keystrokes; opinions that quite literally aid in the needless death and suffering of others while also just generally being a complete waste of time.
In the context of a claimed difference of "several orders of magnitude" (i.e. >1000x), a difference of 1-2x rounds down to 1.
If you follow the citations for that Wikipedia claim (a very very basic step) you will see that they don't provide any evidence for an order of magnitudes difference.
I'm not usually in the habit of shilling for Big Pharma, but since you've taken such a condescending tone I'll make an exception and point out that according to that exact Wikipedia article you linked, literally the same paragraph on the topic mentions that the majority of marketing spending is on free samples. Unless you think that drug companies should raise their prices in order to fund more R&D (which would have the same effect as reducing free samples) this doesn't seem very objectionable.
Just my opinion, but I don't think "private markets" are incompatible with a proper valuation of human life?
It's a bit macabre, but, to me, "dead people don't pay taxes / generate GDP" seems pretty interchangeable with "dead people don't pay insurance premiums / contribute to a risk pool".
To me, the problem is that billed medical costs (what the hospital charges you) have come unhinged from realized medical costs (what the hospital spends on helping you, including salaries, supplies, etc).
If that world is alien-sounding, it's similar to how the billed education costs at universities have come unhinged from realized education costs. In both of those environments, the ballooning costs seem to be propped up by well-intentioned but fundamentally "unnatural" source of funds, in the form of government subsidies.
These have been exploited by profiteering and accounting optimizations, over the course of literal decades. If your surgery costs 2k to perform, the hospital will happily charge you 20k, collect 3k in insurance, and write off the remaining 17k as un-recovered medical expense, for which the government will reimburse some portion.
The more obvious solution to me, would be to take a hard look at how these subsidies are actually used (/gamed), and steadily reduce/focus them over time, so that hospitals don't have an incentive to inflate costs. The source of the problem isn't anything fundamental about "the private market", it's rather about the current set of public incentives that have distorted the private market into it's current form.
So, this is why, at least in the US, EU and a few other countries, we have orphan drug policies.
If you don't know orphan drugs are drugs that could be a life saving medical necessity for rare diseases that do not have a large enough base of users to make the drugs affordable. I'm not talking about Shkreli "affordable", but it's literally not feasible to spin up manufacturing, or final trials, or educational materials for doctors affordable. So governments offer massive subsidies or cost investments for some of these drugs that often only help TENS of people, not hundreds, or thousands or millions.
What's missing is a similar policy for funding otherwise for-profit efforts like the authors.Educational tools and products are often seen as less valuable but I believe they could be immensely helpful.
Now as to the authors original flaws? Well, his product has a clear sell-through referral potential, but it's not really functional to do so with prescription drugs. That being said, selling this product to a CVS or a webmd as a feature could be really good, or major hospital networks, or heck even pharma companies. But all of that is often year+ sales cycles, you can't just luck into that.
Universal healthcare won't help you there, since it also requires a stringent cost-benefit-calculation, e.g. the British system is very (in)famous for this.
That is for approving new treatments, not for ongoing lifesaving care, so is an apples to oranges comparison IMHO.
The drug needs to have a proven benefit, which usually comes down to how much longer does it give you, if it's only marginally better than the existing treatment, but orders of magnitude more expensive they work with the manufacturer to bring the cost down to where it can be approved.
A lot of what was written about the UK system during the ACA debate was outright falshoods or misrepresentations (e.g. the infamous "death panels", whilst ignoring the Kafkaesque nightmare that can be the US system).
Tell you what, I'd take the NHS over what you've got any day of the week, as would the vast majority of people in the UK.
There isn't a "british" system because health is devolved. There are four separate systems, one each for Scotland, Northern Ireland, Wales, and England.
> requires a stringent cost-benefit-calculation
There are lots of problems with the English NHS, but this is almost never one of them. Unless you have particular examples?
Ah, but then you have all these with the ever-increasing demand for profit and the issues caused by that incentive on top of it all. Is that an improvement?
An honest question, what value does the government assign to the human life in your opinion?
In my experience the elites from countries with universal healthcare seek treatment for themselves and their families in countries which have private options, while the majority of the population is stuck in long waiting lines and substandard care.
In Micheal Moore's Sicko Slovenia was mentioned as role model because its universal healthcare. Well it just happens that if you need a thumb operation you only have to wait 13 years there to get your turn. [1]
Wouldn't the rational response be to asses both pluses and minuses of public and private healthcare, recognize that not all health related problems are the same and build a better system from first principles rather than shoot blanket ideological responses?
At least in the above scenario the hospital has a financial incentive to treat you (part of why we perform loads of unnecessary surgeries in this country).
I’m for a public system I just think it is very difficult to do correctly, and if we go into it with idealism instead of extreme pessimism we will fail to secure it against those who prey on exploiting public money.
What is the proper value of human life? Does it change over time?
And is the budget-wise national-healthcare-system administrator qualified to make that determination for you?
And, if they do not want you to get better while waiting on line (most people do bet better for most things, except, broken legs, without treatment) or want you to die while waiting on line, why are the lines so long?
In any case, even in Singapore, with arguably the best-balanced system in the world, things are complicated, when people's lives are on the line...
No it isn't. Even in the UK the much vaunted GP will reject any health initiatives unless they help improve their bottom line. And the initiatives are from the NHS and are free!
> But the key insight is that you have to take a step back and just treat healthcare like any other for-profit business. Doctors and hospitals care about what generates revenue. Insurance companies want to save money. Pharma companies are looking to advertise to new customers. It's no different than other industries.
This is one of the most concise and effective arguments I have read for Medicare-For-All, Single Payer, etc.
I'm all for removing the broken system we have today, but a single payer system doesn't remove the challenge that the entrepreneur faces - it just changes it. Under single payer, you are now trying to sell to a large government organization, with all the pain and red tape that comes with that.
Not necessarily. Doctors can still have individual discretion over their choice of IT service provision (etc) under single payer.
The only thing single payer does is eliminate insurance company opportunism. It doesn't suddenly convert the entire healthcare system into a dystopian concrete monolith.
Although if it did it would make OP's job far easier, because if you could prove there were cost and health outcome benefits on a national scale you would be able to convert the idea into a viable business with a single corporate sales pitch, instead of trying to sell to every doctor individually.
In fact the real resistance would come from pharma. Single payer or not, evidence-based prescribing of empirically proven best-in-class drugs would undermine pharma's entire business model.
> because if you could prove there were cost and health outcome benefits on a national scale you would be able to convert the idea into a viable business with a single corporate sales pitch, instead of trying to sell to every doctor individually.
As someone who ran a healthcare business in Canada, I can tell you that this is absolutely not the case. Single payers are outrageously conservative buyers... sure, you may convince them in 5-10 years, but no startup can spend 5-10 years burning cash without making revenue. Particularly since, unlike with drug discovery, there's no guarantee that you'll be the only one selling this if the single payer eventually approves it.
What actually happens is that the single payer orients itself to big, slow companies that more closely resemble it, thus devastating small businesses and rewarding large, established players. It is damn near impossible to create a startup that sells to a single payer healthcare system.
That's a shame, in the UK we've seen some medical companies do very well by selling to GP surgeries and Primary Care Services (e.g. Egton who provide the online prescription and appointment booking service PatientAccess).
Providers are still basically private businesses (at least in primary care), so it is possible to sell to them. The problem is that each province runs its own system, unlike the UK where the NHS is universal across the country. So your addressable market is limited unless you can tailor your product to fit all 10 provinces. In practice it's very difficult, which is why almost all of the electronic health record software is owned by one of 3 large companies.
> As someone who ran a healthcare business in Canada, I can tell you that this is absolutely not the case.
Correct me if I'm wrong, but doesn't Canada have publicly funded providers? I'm all for single payer, but with a completely privatized provider network. Make the free market compete for the government's dollars. Technology should theoretically reduce costs because it more automation generally means higher profits.
You are wrong. The single payer in Canada is each province's respective health insurance plan. Anything listed on their "schedule of benefits" can only be paid for by the province.
Providers are actually in many cases like contractors... most family doctors, for example, bill for services rendered and are otherwise free agents (unlike in the UK).
Presumably the government will be incentivized to try to save lives in a way that simply doesn't exist for doctors and insurance companies trying to make a profit.
What happens in reality in systems like that is that doctors are lectured to avoid prescribing expensive exams, things like MRI or PET scans, even sometimes not telling patients they should do them if they could afford to pay them on their own.
At risk of identifying myself, I work in this industry (or should I say, I work in the medical field) and yes... it's insane. And of course, you get that privileged peak behind the veil, you get to see how the sausage is made as it were.
It is truly humbling because there is no easy solution, there are no unified APIs; you couldn't use them even if there were, because your medical records are highly sensitive. So forget about poking around production to solve a bug, you've got to do it through observability.
You can't move fast and break things because your downtime will get the regulators involved, because peoples livelihoods are at stake. You will get official complaints through various official bodies who can and will sanction the business.
In short, it's not a fucking joke and it takes a lot more time and a lot more money to do something serious here, and for good reason.
Note that I'm not talking about low-key wearables like Fitbits that monitor your pulse, I'm talking about doctors, triage, etc.
>I think people are being a little too hard on the author. Startups in healthcare are hard.
They are hard, but this case isn't really healthcare-related. He tried pitching his product to practices without showing how it could increase their revenue, or decrease their costs. Without that, his product is just another cost-center with some marginal (and ambiguous) healthcare benefits.
This. Been through two medical device companies that pivoted from clinical trials to product and the major hardships came from not fully understanding the money trail. In the UK the 'Cathedral' of the NHS means that the benefits may not be felt by the your customer.
I still think this idea is powerful. Not because I believe it transforms healthcare, but because it informs consumers a bit more. That’s what seemed to get you excited in the first place.
There are a few paths I see still:
a. Assuming you own a database, expose your dataset as an API. Consider possible consumers. Open source it, and monitor the usage. Who’s using that data.
b. Cold message a few folks on Linkedin from pharmaceuticals that your algo says has high clinical success. Can you “recommend” them? Do they see value. Maybe you have a research tool on your hands.
3. If you can keep your website alive, focus on SEO and gain traction with more users and play that out. Write blog posts, use affiliate marketing, and grow users. But don’t add new features. Just maintain what you have. Try that for a year at least. What did you learn?
I think the toughest thing here is knowing when to stop building the product. That’s what stands out to me as your demise.
The doctor was spot on, and those are the kinds of things you want to hear. It just means looks somewhere else. Writing a doc beforehand with potential opportunities is always a good idea, but it’s never too late.
Disclaimer: I do product at MEGACORP and love product strategy.
This is a classic case of "it's not a product, it's a feature". This is true of almost everything in healthcare that's software-related: it's only useful if it's integrated with the EMR. This is why you never see software startups in healthcare.
In this case, similar functionality (drug recommendations) exist within EMR systems. Except the recommendations are issued by other doctors and medical researchers, not some ML algorithm and it works transparently with the pharmacy management modules.
Moral of the story: don't get involved with healthcare startups unless you know the structure and economics of the industry. They are very different from consumer / business tech.
There's barely anything more frustrating than a fascinating discussion on HN and then folks start kicking around lingo and acronyms that you can't even Google because they could mean a million things.
So please, have mercy and cure my ignorance and explain what EMR means.
EMR is basically anything and everything related to the operation of a doctor's office and/or hospital. Think scheduling, patient records, ordering lab tests, sending prescriptions to the pharmacy, billing.... all of it. It's what your doctor is furiously typing data in to during your visit. They also include a lot of functionality designed to, among other things, make sure your doctor doesn't write you prescriptions for a drug cocktail that will kill you.
It's spoken about as a monolith because that's how it's sold -- there are maybe a dozen platforms in wide use (which are then usually customized by a system integrator for a big hospital). They do have every option you could possibly think of.
I'm not criticizing you here (you're not trying to do a medical startup), but anyone who is trying to do a medical startup needs to know a whole lot about EMR systems -- more specifically how they are bought and sold -- if they want to make any traction. All medical software flows through the EMR, and they are closed platforms by design.
I was thinking the same, except OP's product seems more like a feature for an EMR platform and less a standalone product.
The only downside is that sharing this information may open up the hospital to more liability - i.e. a patient takes an opiate instead of an NSAID and then blames the hospital for recommending the opiate to them because of how much more effective the opiate is at treating pain in clinical trials.
Maybe you could open source it, and with the help of people from the community get sort of grant from Google or Amazon or say digital ocean so that this could work without you having to pay for servers? All the work that contractors were doing, people could do that for free. The only thing is I’d license this thing so that it can’t be used commercially (except for research)
Another possibility: how many people would be willing to donate a few bucks a month to support it? Call it the public radio model. You won't get rich, but you might pay for upkeep and servers.
Donations are extremely difficult to make work. I tried it after getting decent traction in 2009 and didn't receive a single donation to cover server expenses. I'm guessing with Patreon things have gotten a bit easier but for a business I would highly advise against going down the donations route. In general, small payments also tend to take up a disproportionate amount of your support time compared to larger transactions so you have to be prepared to deal w/ that.
One big difference post-Patreon is that I no longer think "my small regular contribution is just going to get chewed up in transaction fees", since Patreon does just one transaction a month and then divides it up among several people/organizations that I support.
But, it would not be my first business model to try, just an alternative to "we built it but now let's just shut it off".
Biggest rec. is to talk to people who have done something interesting or catches your eye. I always email people asking how they did something - people usually are really friendly and willing to share.
Websites like Indiehacker, stratchery are good. Would also recommend the innovator's dilemma if you're at a big company. There are also a few videos on Youtube by CNBC on products of the past - and I've enjoyed those.
I'm so confused by the tone of this article. Obviously some of the jokes are parodying startup culture and are meant to be funny, however I think the "fantastic idea" stuff is meant seriously?
"So I built something people wanted. Consumers wanted it, doctors wanted it, I wanted it. Where did I go wrong?"
As far as I can tell, nobody wanted it except the author. The one doctor he talked to didn't, and no consumer ever saw it. I feel like the author learned a lesson, but not the full lesson.
Reading this was like watching someone you don't politically agree with doing comedy. You know they're trying to be funny, but you also know they're missing the whole point and aren't self-aware.
The funny part is, he might still be wrong about it being worthless. He talked to ten doctors? There are 1M in the US. It's kind of funny to drop all that time and 40k in then give up after ten doctors say no, especially without trying to tweak the product based on feedback.
I, as a consumer, find it worthless and they shouldn't have even started it, but now that you're here, might as well give it a bit more of shot than that. To completely bail on what you have after 10 no's is also the wrong move.
Doesn't seem like he asked the question, "well, what would you pay for?"
He should obviously be selling to clinics / hospitals
Plus when did the 'we don't have any budget' excuse stop a SaaS company? That's what you raise VC for - build the product, hand it to the customer for 'free' and charge them what it was really worth based on their usage when you threaten to take it away
Or try insurance actuaries. If _better outcomes_ is the goal then doctors will want to keep their premiums low by prescribing the optimal drug. If the actuaries are able to see what the doctors are prescribing and what is the better fit they might be able to increase premiums for doctors that take more risks prescribing off of the meta-analysis.
Or, pharma. They have pockets and imagine what that graphic could be worth to Aleve. Otherwise reps could use this data in their sales pitches. Lots of money here.
Have that doctor introduce you to her pharma reps.
Find other data that is specialty specific and sell to pharma in that specialty.
I’m a healthcare CFO and this could work. You’re talking to people that’s a great step.
Keep talking. You built the thing, put it on a $30/month server. I suspect you don’t need the cloud stuff. Just make it your proprietary graphics & viz data. It doesn’t have to be freely available.
Raise money for this. You’re in the Bay Area for a reason.
Get commitments from pharma first. Keep more of your company from PE.
CVS doesn't care if you buy Tylenol or Advil or which one works better, and makes no additional money depending on your choice of product. In fact, the brands pay CVS for the privilege of being featured on an eye-level shelf vs at the bottom.
I think it was a great article. Like it said: To succeed, an offering must create value for all entities involved in the exchange—target customers, the company, and its collaborators. And value is literally measured in dollars, so you're either saving or making money for someone, otherwise it's just a fun project.
I would absolutely choose one store over another if they had a proprietary tool to tell me which medicine is scientifically most likely to work for my ailment. I cannot stand the massive wall of choices at pharmacies.
This is a pretty short-sighted way of thinking of it.
A tool like this could easily be marketed as a USP for a pharmacy:
"Only CVS has MedFinder: a proprietary system to find you the medication that's most likely to help cure YOUR SPECIFIC SYMPTOMS. No more standing in front of the racks of too-similar sounding medications. Enter your symptoms, get a recommendation backed with data!"
The product here is a kiosk backed with some real data that you can build a marketing effort around as yet another way to try to differentiate yourself in a heavily commoditized business. Pharmacies fight tooth and nail to get people in their door, and especially, to transfer prescriptions to them. One more small detail that you can convince people might help them, that actually gets them through the door? You could build an entire campaign around it.
(Think Dr. Scholls placing those foot-pressure-sensor systems in pharmacies. They drive, pun intended, foot traffic.)
Is it possible to sell to single payer health providers in other countries? Something like this could have massive implications for, say, the UK's NHS.
Yeah. For investing so much time and money it seems like he gave up after a couple of bad meetings.
It takes years to figure out how to properly sell your product. He didn't find out that his product wouldn't work in the marketplace he just found out he didn't know how to sell it.
> Reading this was like watching someone you don't politically agree with doing comedy. You know they're trying to be funny, but you also know they're missing the whole point and aren't self-aware.
are you not supposed to laugh at the jokes of someone you disagree with politically? is it possible someone could be "self-aware" and still disagree with you?
I take the article to be written in the mindset he had at the time. When he made that quote you mentioned, he thought as he was heads-down building the product that it would be useful to those people.
He then goes on to describe the evidence specifically why doctor's DIDN'T want it. All the author was trying to do with that quote was explain what he was thinking at the time.
Ya, I'm surprised it's not just a static landing page for consumers with ads. At .50 cents per user, that's only 80k users required to break even (ignoring the time variable there)
I'm no adtech expert, but I'm pretty sure he would not be getting .50 cents per user per year - that's what WebMD, one of the top 1000 most trafficked web sites in the world makes, so I assume it can sell it's ad space at a much higher price than a greenfield site with no traffic would.
The first example the author gave with painkillers is inherently bad.
Sure, Aleve is the most powerful. But it also has the most side effects, including ulcers if used for too long. There's a lot of nuanced information when it comes to drugs. If you want to go on pain killers like Aleve for long term, you probably need to take acid blockers as well, like Nexium. So where does that fall into the analysis?
Unfortunately a lot of stuff requires medical doctors that have experience. I just went through a severe health emergency, where I was taking one antibiotic and needed to be changed to another. I needed blood tests and constant supervision for the doctor to make this decision.
Psychiatry, which is what I guess the author targeted, is likely even less reliable. My friend is a pharmacist and was on anti-anxiety medication for a year, but decided to try another one, because of the effects on her were affecting her and now she's doing much better. This requires working with a psychiatrist for a year.
I think there's value in surfacing all this information but to expect people to rush into paying for it is naive. It's a nice-to-have-tool that might be useful like a Google-type research tool but nothing someone would want to pay $5/month for since the actual benefit over a psychiatrist just going off experience is completely unknown. There no inherent home run here.
>Sure, Aleve is the most powerful. But it also has the most side effects, including ulcers if used for too long. There's a lot of nuanced information when it comes to drugs. If you want to go on pain killers like Aleve for long term, you probably need to take acid blockers as well, like Nexium. So where does that fall into the analysis?
Am I wrong or isn't solving these issues basically the service UpToDate offers?
I wish he could have invested more in the idea - I think it's an amazing concept but the creator pulled the plug too early. Sales cycles in B2B can take months, and in healthcare they can take years.
Especially the doctor's feedback, she described the possibility of changing her habits slowly as she got used to using the tool.
Unrelated:
"I always have the best interests of my patients in mind, but, you know, it’s not like they’ll pay more if I prescribe Lexapro instead of Zoloft. They won’t come back more often or refer more friends. So I’d sorta just be, like, donating this money if I paid you for this thing, right?”
This line kills me and it's exactly why I hate the healthcare industry. She clearly understands that the software could improve patient outcomes, but describes it as a donation and still wouldn't change how she writes prescriptions (out of laziness, I guess?)
Sounds like there's an opportunity to have insurance companies pay for the product. This would directly align incentives and lead to overall cheaper care for the patients.
I had the same thought. The meta-customers here are the insurance companies. If you can prove that the tool improves the performance of doctors so that they get sued less when they use it then the insurance company will give discounts for doctors on their insurance for using it. Presto! We just created a market for doctors to pay for the service.
That anecdotal interaction is extremely damning for the entire US healthcare system. Physician admits she prioritizes practice revenues over patient outcomes. Other physicians express similar preference. Inventor of useful tool tosses it into the mothballs.
That tool isn't just a physician aid. If that's what doctors are really doing, it's more like a physician replacement. If they won't do evidence-based medicine, and prefer to be legal drug dealers, for profitability reasons, we should all be contributing to the creation of the auto-doc, instead of trying to help them.
That $40k would have been better spent lobbying to remove the guild protections from the practice of medicine.
If all general-practice docs do is take height, weight, pulse, BP, blood oxidation, and run a CBC with lipid panel before prescribing your drugs for the year, that doesn't seem like something only an MD or NP could do. But by law, if you want prescription-only medicines, you have to go to a prescription-drug dealer and pay them for the scrip, before going to pay someone else, again, for the actual drug.
And if they're not actually doing (what I perceive to be) their jobs? If they're just "phoning it in" to collect a paycheck? That's one job I'd love to automate out of existence.
I don't know how old you are, but at a certain point, the input of a general practice doctor becomes very valuable. Remember, the doctor doesn't even do the little tests you mentioned, like taking measurements and blood tests - they just order them, and nurses or technicians do the work. The doctor's job is to make decisions and give advice.
About a dozen years ago, my doctor caught my blood pressure shooting up in my annual exam and started treating me for high blood pressure. High blood pressure is a "silent killer" that leads to premature deaths for millions. You don't feel sick, while your heart struggles and your brain risks explosions. And treating it, once diagnosed, isn't straightforward. It took six months of experimenting with various medications to find the combination that worked for me. That's a doctor's judgment and expertise, not something for some technician with a community college associate's degree. And this summer, when a critical potassium depletion drove me to a scary ER visit, she noted that one of my BP meds could affect potassium and adjusted my dosages.
And it's not just an expert-system thing that could be replaced by a robot, either. She's not just dealing with my measurements - she's dealing with my fears, and helping me decide on best approaches. That requires a human touch.
So your doc abrogated her responsibility to warn you about the side-effects of the prescribed drug.
You went to the ER, because you were not aware that cramps, constipation, spiking BP, and muscle weakness are symptoms of hypokalemia, which could be resolved quickly by dissolving KCl from the grocery store in water and drinking it.
That's patient education that needs to happen before you go to the ER.
Just like with the GlacialMD thing, it's no skin off your doc's nose if you have to drop $500-$2000 on an ER copay because you had an electrolyte crisis from the medication she prescribed, and didn't know ahead of time that could be a possibility.
And the rest of you out there, learn to measure your own BP. You can't leave it to an annual exam to find out it's getting high, and the BP taken in the doc's office may include the "white coat effect" (or as I call it, the "what will this cost me effect").
To be fair, I took that particular medication for over ten years before the crisis. So it's unlikely that the medication itself was the sole cause.
And when the symptoms started, I wasn't going to go to the chemist to start concocting home remedies for what I feared could be a mild heart attack. I went to the ER. Again, I don't know how old you are, but I'm old enough to have had friends my own age start dropping dead of heart attacks and strokes. Hence I decided to take the safest route.
I agree with you, but at least she was honest. Many others would say no for the same reason, but wouldn't be honest about it. A prospect who is honest about their true reasons for saying no is extremely valuable when you're pitching an early product.
I think in general, if you're building a business, expecting anyone to do something that isn't in their own selfish interest in some way is a recipe for failure, even if people in your target market might claim not to have selfish motivations. If you want someone to change their behavior (and/or their organization's behavior), you need to be able to clearly explain why it's a win for that specific person.
"Laziness" is probably unfair - but there is a lot of evidence that everyone, doctors not included, will get stuck in their ways unless forced out of them.
There's much to be said for sticking with things we know work, rather than experimenting with new things that might work better. If a doctor has an effective treatment for a common condition among their patients, what is the cost of trying something new? Do you try it with all patients, or just one?
But "laziness" is something of a trigger word for me in health care discussions, just like "greed" is. It's so easy to attribute what we see as suboptimal outcomes to malice on the part of the person acting - even passive malice like laziness.
Congratulations on learning the hardest lesson, your next startup is now a lot more likely to succeed, so get to work!
A lot of comments criticizing what the author already knows, that the business plan was incomplete. I had the urge to do the same, but read to the very end and realized TJC clearly learned the most important lesson, and it's one I've learned myself, so I have no room to be critical.
BTW, I burned a lot more than $40k and 9 months in my own story, so maybe it helps to hear that other people have made the same mistake and gotten too convinced their idea is valuable without checking carefully enough -- without quite knowing how to check carefully. I know what it feels like to know that I'm doing the right thing, and then find out that it's not good enough.
Fun read, and an important lesson that many very bright people will still learn the hard way, even after reading a story like this. Thank you for sharing.
Realistically, while I feel that the author is on the right track with this data driven approach to evidence based medicine, the fact remains that the advantages/disadvantages between drugs within the same class of medicine tend to be negligible compared to drug costs. There are then other specific concerns that must be considered eg pregnancy/breastfeeding, liver failure, kidney failure and often what ends up is that one defaults to a standard drug per drug class with alternatives in specific circumstances.
The real benefit IMO would be marketing this to countries with a top down healthcare purchasing system (eg the UK) or as a tool to drug companies/researchers looking to make their own meta analyses (eg abstrackr). It might also be better instead to diversify into correlating symptom clusters with diseases because then the utility to the end user is bigger.
Sell this product to insurance companies so they can recommend cheaper/safer drugs to their customers, or show the more expensive one and if it truly had better outcomes. Everyone benefits from that.
That said, the medical establishment is a tough tough domain to move.
Insurance companies already have teams of doctors and pharmacists analyzing studies to ensure they are pursuing the most effective and cost effective healthcare. They’re called Pharmacy and Therapeutics (P&T) Committeea.
I worked at a company in the early 00s that did exactly this -- natural language processing of medical records for meta analysis. Even when it worked, we couldn't sell it for much. The budget of each study was too low to really build a business around, and the low number of potential buyers didn't help things.
Eventually we pivoted the tech to a B2B SaaS play because it was too hard to sell.
The author says that they fired the contractors and closed the site. I can never relate to this. You built something that is at least somewhat useful to at least some people. Why not just keep it online forever? It can't cost more than $5/mo to host this, I have a server where I host around ten of my projects for less than $20/mo.
Why do people shut things down instead of just leaving them there, unmaintained? Even that is better than just taking it offline.
Well in the case of this particular product that offers medical advice, the info may become outdated fairly quickly. I suspect there is some amount of liability associated with offering potentially out of date medical advice that isn't worth the tiny revenue stream.
Anyone could theoretically get sued for anything. But the chance of getting sued for providing an out of date meta-analysis is indistinguishable from the background chance of getting sued for getting out of bed in the morning. (assuming you're not doing something stupid)
I was tempted to post this same question. I suspect the answer is that the OP just doesn't know how to scale it down; it was built by contractors and he/she might not even have the keys to the boxes, let alone know how to keep them running.
It is unfortunate. Even if the data goes stale and the product is no longer fit for purpose, it shows well and would make a great demonstration for his/her future endeavors. And you never know who might stumble across it in the future and want to revive the technology.
I've had fairly sophisticated sites running on Google App Engine for years after the company was "shut down". One ended up being instrumental as a demo when meeting a later cofounder.
I have a family member who worked somewhere that provided tools to medical practitioners to help them determine how a patient would respond to a drug based on their genetics. They also had a hard time selling directly to clinicians, because as the author notes, they do not particularly care about patient outcomes in the abstract sense, because much like the programmer bravado, doctors have a similar one: most of the other doctors are worse than me, they need the tools but I don’t because I have a more nuanced view, I know what works for my patients better than any statistical study could tell me, etc.
The only way to make money is to get insurers to require these things, or to sell to some big hospital network administrator who you can focus your sales efforts on. I don’t think the product here was mature enough to sign a multimillion dollar contract with a hospital network, and maybe not for insurers either.
> Make something people want. It’s Y-Combinator’s motto and a maxim of aspiring internet entrepreneurs. The idea is that if you build something truly awesome, you’ll figure out a way to make some money off of it.
Nice narrative style and congratulations on achieving your (first?) failure without too much trauma.
It's possible you invested your own money prematurely, possibly without talking to enough people about the opportunity. Your metric for how much people want something is usually directly tied to revenue, or if you are using engagement as your metric, you should have figured out some way to make money off of it before you invested in development. I suggest that other engineering-minded founders think of sales like writing "hello world" -- for example, if you're doing a physical consumer product, you can start by shipping pet rocks on day one.
More than just identifying a product people want, you also want to find a network effect or viral marketing approach so you can grow with a sustainable advantage.
I think the original poster made the same mistake I did, and took the wrong lesson from the situation.
The startup idea sounds great and I would enthusiastically start working on it right now as a cofounder.
The problem was spending money before having a sense of how to get more.
That’s all.
If the startup could have been run entirely without paying contractors or incurring other serious expenses, the poster could have run it for longer, and iterated on models longer, until the product/market fit got dialed in.
There just wasn’t enough runway, enough iteration.
Oof. The site was practically named after WebMD, and the founder didn't do any market research on that competitor to gauge how viable the business was before quitting their job? This person probably should have known that their main competitor made $0.50/user before they quit their job. Also:
>It had been a bit of a working assumption of mine over the past few weeks that if you could improve the health of the patients then, you know, the doctors or the hospitals or whatever would pay for that.
This person uses the "X% of Americans" line, so they live in the United States, right? The one in North America that views healthcare as a business to wring money out of? They definitely should have done more market research if they could type that with a straight face.
But hey, hindsight is 2020 and they only spent $40k on some very valuable firsthand experience. Sounds like a win if they're planning to run businesses in the future, but for ideas without a clear path to making enough money to sustain themselves, try to use other peoples' money next time :)
That line stood out to me also. Anyone who feels that the primary goal of healthcare in the US is to improve health outcomes is incredibly naive about how this industry (which is over 20% of the us economy) works. There is only one group in the entire industry which has an incentive to keep people healthy (insurance) and that is the most hated of the bunch!
Perverse incentive structures are the fundamental reason US healthcare is so expensive and stupid. It's not "greed" in a simple sense, and it's not "regulation" in a simple sense. It's a system of interacting problems that needs treated in a systemic way.
Blaming the insurance companies strikes me as the political equivalent of yelling at the cashier.
Insurance is one of the oldest industries in the US, which means they have accumulated many (probably well meaning, but ultimately stupid) regulations over the years.
For example, to be taxed as an insurance company (e.g. avoid paying tax on premiums that will partially be paid out in the future), most states assign a maximum profit margin.
That means, in order to make more money as an insurance company, it's perversely in their best interests to increase the cost of medical care as much as possible. Medical care is not like cat bonds that might wipe out your risk pool, so it's just a cashflow game.
That's not to say there isn't any downward pressure. On a day to day basis, individual claimants eat more into that margin than was budgeted, so a company might have to raise rates higher than their competitors (essentially they have some pressure to deny claims), but there is very little systematic pressure to reduce overall costs. Also, marketing budget does not count towards profit, which is why insurance companies are in this rat race of brand advertising to steal customers from each other. With auto insurance, for example, upwards of 40% of your premium is just going out the door as ad spend.
Anyway, I would assign a lot of blame in this perverse incentive scheme to poor regulations.
To put it in software terms, it's not a bug, it's an architecture flaw. The "poor regulations" aren't individual issues, but rather part of the fabric of a framework that causes perverse incentives. Therefore, the simpleminded conservative/rightwing solution to "reduce regulation" doesn't eliminate the perverse incentives - it just changes the flow, the way tossing a rock in a river does.
If I were the one making the decision, I'd change American healthcare with one big thing, modeled on the Japanese system... price controls. Japan's system is more or less like ours, but with one key difference - a government panel sets the prices for all medical goods and services. Providers and insurers have to make their profits within that price structure. This incentivizes both to reduce costs, for the sake of their own profits.
Of course, this would be called radical socialism or some such stupid in America, and Democrats aren't even talking about it (Republicans have no ideas at all on health care). But it's politically a much lower threshold than single-payer (throwing out the whole framework and starting over, basically). And it demonstrably works - Japan's system costs less than half what ours does per capita, with better outcomes.
> There is only one group in the entire industry which has an incentive to keep people healthy (insurance)
The motivation of insurance is to keep the cost of treatment less than the income from premiums -- it doesn't have anything to do with how healthy people are.
Yep - the motivation of health insurance is to keep people out of hospitals and away from healthcare professionals as much as possible.
Sometimes that aligns with keeping people healthy, but other times that aligns with women not getting routine prenatal screening and thus increasing the infant mortality rate.
.. and the doctors he talked to felt they already had the knowledge they needed! If anything, they felt, perhaps, that this devalued their existing knowledge. And they don't make 10% more money for providing 10% better outcomes.
This is probably the root of the problems in US medicine: there is no effective feedback. The ACA (Obamacare) did a little work on this with Medicare payments to hospitals, but that is an exception in the system.
What could we do to make this the basis of the system? Without driving doctors away from patients who have bad prognocices?
Take a look at HMOs, such as Kaiser Permanente. HMOs combine the function of insurers and medical providers in a way that aligns incentives (everyone pays about the same, but healthier people need less care and are thus more profitable customers). There are many problems that come with the HMO model (e.g. the potential for low quality care), but when it works, it directly solves the problem you addressed and exists in the US.
Imagine you have a chronic condition, and every month your insurance writes a substantial check for your care. A new treatment comes along that might fix your condition permanently. Surely they'll leap to pay for it?
Unfortunately, they probably won't want to. People frequently change insurance, so the most likely outcome is they'll have to pay for an expensive treatment, but never see the savings.
This is currently playing out for some new expensive new gene therapies.
Sure, but the decisions get made at the scale of a single company, not the industry as a whole.
For the mentioned new gene therapy treatments, people have been asking congress to come up with a new model for how to pay for treatment. Here's a recent opinion piece by a Louisiana Senator where he lists off the proposals he's heard: https://www.statnews.com/2019/06/12/paying-for-coming-genera...
> There is only one group in the entire industry which has an incentive to keep people healthy (insurance) and that is the most hated of the bunch!
A testament that the demand for healthcare services is not only about health results. Albeit there are grave challenges about the way Kaiser or HMO's have supposedly better health results.
This is changing. Medicare, medicaid and to an extent tricare are switching away from pay for service models, so there is a financial incentive for providers to improve health outcomes.
At Healthify we’ve built a business by aligning positive health care outcomes with the incentives of people paying for healthcare. I think it’s incorrect to say that American healthcare isn’t interested in positive outcomes, it’s just that the systems in place have made it difficult to track and understand those outcomes in the first place.
We just raised a series B and are hiring engineers (very remote friendly)!
OK, so build a site that takes a bunch of patient info (demographic, genetic) and generate 1000s of custom meta-surveys that estimate the most effective medication for that specific patient by weighting the data in various ways. Return only the one that suggests the highest-priced medication. Eventually, move to a bidding system where pharma companies can sponsor the custom reports that recommend their products. Give 5% of the profit to developing world vaccination projects and it'll save lots of lives overall.
IMHO the point of the article is its title - and certainly, furthermore, the point made at the end:
“To succeed, an offering must create value for all entities involved in the exchange—target customers, the company, and its collaborators.”
I believe the point of this comment is 'do market research before quitting your job'. The author could have learned their presented points while operating this as a side project concurrent to their existing job.
> they only spent $40k on some very valuable firsthand experience.
This. It is very hard to get real world experience. $40k is cheap compared to a bootcamp that won't teach him even 1/10th of this life experience.
Btw, actual figure is not $40k, but one should consider it close to $100k+ if you factor in that he used up his "make my parents support shelter, food, living expenses, car". Depending on the cultural background, they wont be doing this many more times. If you consider how much he lost on the (bay area) salary he otherwise would have earned, add another $100k
Yup. Depending on how attractive OP is to high-salary big tech employers, the opportunity cost alone of not getting paid for most of a year could easily be several hundred thousand dollars.
In the story she describes the project as "fun" before he asks what she'd pay. At that point he probably should have picked up that she was being kind and was not in fact seeing a product she was interested in buying.
He probably could have figured that out by asking doctors before he built the thing.
I think he should target not WebMD audience, WebMD doesn't provide searchable and visualizable database of trials and outcomes, but more professional crowd: research institutions and universities, government, pharma corps, pharmacists, hospitals, and charge some subscription fee (e.g. $200/m), and it would be totally different story.
You spend the whole comment putting down the author for making mistakes and then try to make it look like you're retracting this by putting a positive spin on it towards the end. Not because you especially care, but because you don't want people to downvote you to shit for this obvious trash talking.
"try to use other peoples' money next time (smiley face)" A++++ douchery. Nicely done.
The most disappointing thing in this story is that glaciermd.com doesn't even resolve anymore.
He started a business in 2017, and closed the doors in 2018. What?
Recently, I was contacted by a person who told me he saw one of my videos on YouTube about Ycombinator (it had 50 views!!!), he had to meet me in person. We met a couple days later. He was a very smart guy with an impressive resume.
He quit his job in equipment manufacturing to pursue entrepreneurship. He was a former USC student, so naturally he traveled back to California, rented a small room where him and his business partner churned on their idea. It was a sort of classified for colleges.
Not a terrible idea in anyway. He knew the college life so he must have known things I didn't. I told him he should have tested his idea before quitting his job and moving into a whole other country where he had only been a student. He showed me the website, and it was obviously made by someone who was learning React. There is no need to go fancy when you are still testing your ideas. But there was a problem.
They were paying thousands of dollars a month to keep this barely functional app running.
This is something I see a lot of young people entering web development are doing. They get beefed-up cloud services that eat up a good chunk of their money. Most people think this is the regular cost of being online. It's not! A year later, the domain name doesn't resolve either.
In the story, OP alludes living with his parents. I imagine to save money. But then there is $40k. I hope that it wasn't mostly on the AWS cost. But he gathered valuable data that did cost money.
Don't shut down the website. Keep it running and make the data available in read only mode and add a donation button (alternative to ads). It should cost $5 a month to host on linode or digitalocean. Don't let the money you spent go down the drain.
I'm so confused. Not like I'm a doctor, but doesn't the foundational assumption seem fatally flawed? How can you reduce drug efficacy to a single number? It's one thing to compare brands of the same drug, but why in the world are aspirin and acetaminophen (or whatever) just viewed like interchangeable substitutes here? I know I certainly don't reach for the strongest painkiller—but rather the weakest one that might do the job. And I know my doctors certainly never viewed these painkillers as interchangeable. Don't the natures of the effects and side effects actually matter to people?
> “And that isn’t worth something? Prescribing better treatments?”
> “Hmmmm,” she said, picking at her fingernails. “Not directly. Of course I always have the best interests of my patients in mind, but, you know, it’s not like they’ll pay more if I prescribe Lexapro instead of Zoloft. They won’t come back more often or refer more friends. So I’d sorta just be, like, donating this money if I paid you for this thing, right?”
I've never seen such a concise indictment of the American healthcare model.
If this thing works as a better search engine for medicine, then I'd want my doctor using it. Could this person have trialled it in a healthcare system with a public-option?
Convince the government your product is a benefit to their doctors and by extension their citizens, and a good government should pay you for it after a proper trial.
Instead he walked into a situation where the incentives are utterly fucked by an atomised, privatised, and consumer-oriented healthcare system.
While not a unicorn, this could potentially succeed as an indie-hacker sized project. One potential business strategy could be a high-level information hub - a weekly blog + podcast for example. Each week look at a different drug or illness, look at all the relevant papers, do a short statistical analysis, and summarize. Then get some patreons, run ads, maybe sponsorships for particular drugs, etc. I think there's a lot of math geeks out there that would find this useful/interesting enough to follow on a regular basis, even if just to see what's going on in the healthcare space.
Another idea that could make this a viable business: while in general consumers won't pay for this sort of thing, there are certain problems that the medical establishment just does a horrible job treating, where I think people would gladly pay up. Migraines, for example.
I personally suffer from some sort of headache disorder (symptoms are a mix between migraines and cluster headaches, I've had different diagnoses). I went to a neurologist every 3 months for 2 years, at $200 per 30 minute visit, and usually less than 10 of those minutes were spent with the doctor. Every time he made some seat of the pants guess as to what to do next (adjust the dosage I was receiving, or change medications). In the end, I never found anything better than taking a bunch of ibuprofen at the first sign of a headache, and that was something I figured out on my own.
I wouldn't pay a dime to see a comparison of generic and name brand NyQuil, but if a trustworthy source offered me a personalized report on what medications I should be looking at based on my headache symptoms, and had studies to back it up, I'd gladly pay $100+.
This is one of the best startup post-mortems I've read on HN. Clear and engaging writing which also provided key insights into the mind of the entrepreneur during his journey. It's sad that it didn't work out but I think it can help teach valuable lessons to anyone reading it.
It's a shame this project was shuttered as what screenshots are posted look amazing.
Better data aggregation is desperately needed in medicine and sadly I think the creator has walked face first into the horror of cost merely to access clinical trials. Plus, the cost of paying contractors to extract the actually useful information from those trials.
These are key reasons why Open Science and Open Data are so critical, to produce tools like this one. Sadly I don't think there is a "Startup" level of money available for it even so.
It is easy to be an armchair expert, but I saw that this is not a viable idea: almost nobody would pay for it, for 99% of the cases patients and doctors already know the medication that is good enough and good for them (ibuprofen for me, acetaminophen for my friend because Ibuprofen upsets her stomach etc). And remaining 1% will most likely require serious medical attention, not just rando site. Let alone the fact that this is borderline dispensing medical advice which is a highly regulated activity!
One way to look for a similar business is examine.com (disclosure: I do not use them but read their write ups and articles about them on indie hackers-like sites and podcasts). It is a site that gives information about supplements. The supplement landscape is much more wide than that of painkillers, they are hard to compare, and supplements are usually chosen by consumers themselves, and supplement information is unlikely to be treated as a medical advice.
Theoretically I could see this being used as a marketing gimmick for the PCP office (we collect a profile of medications best for you based on a statistical evaluation of results from peer reviewed studies!), and that could help docs make the case that patients should see this clinic vs another one nearby. But anyone who has tried to use healthcare in America knows this isn't really one of the limiting considerations in how you choose a doctor or practice.
Insurance companies have very little/not at all influence over prescribing patterns of physicians. If you could tie this improved outcomes, then there is a business model here.
>Insurance companies have very little/not at all influence over prescribing patterns of physicians.
Walk into any physician office and ask them how many faxes they get everyday from pharmacists requests the Dr. change their prescription. Then ask the pharmacists what triggers them faxing requests to the Dr.'s requesting change in therapies...hint: insurance companies dictate to the pharmacists their recommendations for Docs.
Edit: I should have noted the "tips". When a pharmacist faxes a Doc a recommendation the pharmacy gets a "tip" and if the Doc actually makes the change, then the pharmacy gets a 2nd "tip." In my experience the Doctors have no idea pharmacists are paid by insurance for recommendations and changes.
Insurance companies dictate what treatments they will cover for which conditions.
However, insurance companies are perversely cost conscious, because they are paid as a percentage of medical costs, not as surplus between premiums and costs.
Congratulations on a valuable experience. You have learned a ton and summarized it wisely. This is all worth a lot more than 40k!
Although the typical SF startup drum is about changing the world, it takes a lot more than just a good product to do it, as you've found out, and especially in healthcare, where the barrier to entry is extremely high.
Also, your doctors forgot to mention that Pharma companies actually pay them (indirectly) for prescribing a certain drug. Your software was not just not useful to them, it was actually telling many a story they did not want to hear. Sure, Tylenol might actually be #3 of the list, but their sales rep is a real friendly one, and next time they go out for dinner, it would be a shame to have prescribed something else...
Anyway, now you know. Isn't that an amazing feeling that you've learned so much that others would never dare, and it had only cost you $40k? Keep going, good lessons are worth it!
p.s. Just because your original iteration of the product was not needed by doctor and not valuable enough to be paid by the consumer, it doesn't make it useless.
Perhaps if you made it the #1 on Google, consumers would notice, and use it, then you could use it as a lead-gen for other products, and sell that space/time/data.. whatever.
The funny thing about startup ideas, your first iteration is probably almost always wrong.
Hi Tom, thanks for sharing this. I work for a company in the evidence based medicine space that would have become your competitor had you continued to pursue this idea. Send me a line (email in profile), it would be interesting to get introduced.
Hey is there a way for patients/caregivers to get involved? I'm particularly interested in finding ways to collect data on trials that are in progress but haven't published results yet. Secondarily trying to build an evidence-based approaches to selecting low-risk adjuvant therapies that hasn't been directly proven in clinical trials.
> “And that isn’t worth something? Prescribing better treatments?”
> “Hmmmm,” she said, picking at her fingernails. “Not directly. Of course I always have the best interests of my patients in mind, but, you know, it’s not like they’ll pay more if I prescribe Lexapro instead of Zoloft. They won’t come back more often or refer more friends. So I’d sorta just be, like, donating this money if I paid you for this thing, right?”
So, what's the obvious next step ? Prescribing things that won't work so patients keep coming back but not badenough that they flee to another practice ?
The solution is to educate the patients so that they demand this sort of data driven analysis from their doctors.
If such a product existed I would use it continually and challenge my doctors when they contradict it. If patients aren't educated and care, health professionals won't care enough to purchase it.
The obvious next step (if there is one) is to get the insurance companies to pay for it, since they are the ones who would theoretically reap the benefits.
"Theoretically" is the problem here. Insurance companies won't spend a penny they don't have to, so you need to prove you'll save them more than they pay you starting from day one. That's a tall order.
Its kind of sad picturing the customer with a headache needing a app to tell him what medicine to buy at CVS...while there is a pharmacist right there.
I do like the quip at the end about lacing up all birds and jumping on the bird scooter. This makes me think how badly I need allbirds to join the sharing economy...how great would it be to have allbirds scattered all around your city sidewalks so you have a paid on demand when you need them without having the overhead of actually owning them. Its not sexy and doesn't scale well, ticking all the VC boxes.
Maybe in a small independent pharmacy you have this issue, but biggest pharmacies in the World (CVS, Walgreens, Walmart) the pharmacists will not know the margins of over the counter drugs and have nothing to do with purchasing them.
>Maybe competent but certainly not unbiased.
I suppose they have a bias in the form of legal liability and professional licensure which guide their recommendations. That is, they can be sued for negligence, malpractice and otherwise potentially lose their license to practice. Tech startups are notorious for placing profits before users, and lack accountability in the same way as licensed professionals. In a worst case scenario they fold up shop and move to the next industry to "disrupt".
The reason that you don't get a result for "best painkiller" when you google best painkiller is that there's no such thing.
Why are you in pain? have you had a lot to drink recently? What kind of painkillers do you usually respond to? Why not simply combine them? Are you going to keep track of all of these predicates for each query?
This guy's mistake was not that he mismanaged a company. His mistake was having a shitty idea and being utterly deluded about its value.
Perhaps you gave up too soon, your startup journey is shared by most (all?) startups i.e. where they end up finding PMF is often very different from where they started. It's a matter of sticking with it and if it's worth the opp. cost.
Admittedly, this is Survivorship Bias, since the overwhelming majority of startups fail. As PG said: "startups don't fail, founders give up"
Sure, but that's not the whole story -- the OP would also need to be okay with changing the idea drastically where and when required. As other people have noted in this thread, not much effort was made when talking to the doctors to find out what problem they do have, and no mention is made of ever even showing this product to consumers.
The fatal flaw here was building a product "in secret" and not launching soon enough.
And in the market consumers revealed preference is for $1000/yr treatment over $20K vaccine, or equivalent multiples. When a new vaccine is invented, people balk at the price and complain it shouldn't have been created.
I don't buy this line of thinking - the list of curable diseases/conditions is longer than the incurable ones. Selection bias means the incurable ones get more attention (because they have to be managed). The health care industry is (mostly) not a coordinated cartel; it's a bunch of self-interested organizations looking to get a leg-up on the competition.
Thank you for your vulnerability and sharing your story with such honesty!
There are easy businesses, and there are hard businesses. Your business fell in the latter category, but your tactics to succeed were based on easier businesses. I still think you could have succeeded if you persisted and weren't so eager to get sales and play startup.
I commend your effort. Humanity likely would've progressed more from the application of this data than it would with you working on leveraging autonomous blockchains (not that that's a bad gig either).
I think the general takeaway is:
1. have an idea
2. make some mockups with plain HTML/CSS
3. run it past your target market to see if they'll pay for it
4. if so, build it
This works even better if your target market is willing to fund the development costs.
I still feel like there's an idea somewhere in here – maybe with medicine, or maybe with supplements, nutrition, etc. You could position yourself as a knowledge base and use the data as a marketing funnel for related products. Examine.com has an interesting business model along these lines. (If I was about to start an anti-anxiety medication, I'd sure as hell pay to know which one worked best.)
This is why starting with low or no code is a great idea. He would have saved a lot of money if he went around to potential customers and seen if they would have paid for such a product. Or fake a product with fake data and demo it to see if there's anyone that would pay for the real thing.
I think we mistake these ideas as being bad business because of our expectations. Some ideas take longer to produce returns.
I'm currently working on an idea that has no business plan in the near future. It's a cryptocurrency. I hope to make money from in the future by either selling my stash or providing software and services. But in the meantime, I'm running it as a fun/side-project. The project is picking up steam. I've seen interest growing, some coin trading on an exchange.
I don't expect to make money soon. So I do everything myself. I spend little money on the project, 3-5k per year. I enjoy working on it and learning. It helps me learn not only about my project but also about other cryptocurrency. I think that I will make some money, eventually.
If you don't know how to monetize, you need investors...duh :) Then you can spend some millions working it out! And if you still cant monetize make sure one of your investors is deeply in the hole on this investment and you control the voting shares and make them buy you out.
I can't help but wonder if this would have worked better if the author had:
1) Tried this in a different country. Healthcare in the US is very different from Canada or the EU. Maybe it would fail elsewhere too, but its a completely different marketplace. I have a hard time imagining a Canadian doctor saying "This is better for my patients, sure, but will it help my bottom line?"
2) Worked on a solution that didn't involve so much labour. It was data processing, right? Might be a good fit for ML. (God knows it's being used for everything else.)
That being said, I'm in awe of anyone who goes out on a limb like the author did. He didn't just try an idea, he invested in it. And then he had the courage to tell us about his failure. Bravo!
Seems a case of someone not understanding the economics of the vertical they're entering and not wanting to put the effort (ie: time/money) to learn. Doctors generally don't directly make money by having patients get more healthy. Other entities however do. Insurance companies may, self-insured employers definitely do, and hospital chains may (if they're getting bundled payments for procedures). These entities however have very long sale cycles and are old school (ie: won't listen to some 20 year old). Still, there's a whole massive ecosystem of companies in this space so it's far from impossible. Possible to also partner with one of those companies as a sub-vendor.
There is a value-based angle here. But as the psychiatrist told him, she doesn't make any more or less by prescribing one drug vs. another. If he was able to link the use of a specific drug to improved outcomes (READ: fewer repeat visits, less time in the hospital), then it could be something that ACTUALLY reduces costs.
> On July 2, 2018, GlacierMD powered the world’s largest depression meta-analysis, using data from 846 trials, beating Cipriani’s previous record of 522.
Was this analysis written up somewhere? The author links to Cipriani's study in the Lancet but not to their own.
I don't know that I've encountered such a flagrant display of naivete from someone who actually has an education/proficiency in a profession, who is also apparently well-read enough to engage in such a process. It demonstrates a complete lack of respect for other careers that have a tremendous amount of education requirements and proficiencies of their own, not to mention having a basic social awareness of how people behave and respond in the business world and social realm. I get that he was trying to be humorous, but it wasn't informative enough to make me sympathetic to him at all.
I would apply for an SBIR and try to get some academic support. Maybe the NIH would be interested. Your algorithm and analysis platform might be the real special sauce here. Maybe even license to a pharma company.
My university (University of Hartford, in CT USA) had a required course for all engineering majors: "Engineering Practice". Essentially it was a course on how to take an engineering idea & turn it into a business.
Find a problem, find potential buyers for a solution, figure out what they say they're willing to pay, determine other design constraints, rough design a solution matching the constraints (IE that can be sold for what people will pay), create a business plan, including estimates of how much funding is needed and a method to get that funding, build a prototype, pitch the product to faculty as though they were VCs/banks/other funding sources.
If you're trying to make a product you need to know how it will sell. If you're trying to run a business, you need a business plan! If you can't make money you're better off planning to register as a charity, and your budget will need to take that into account.
This product could maybe have been a good business, if it was aimed at someone willing to pay for the data. Mainly, the insurance companies. They want to make money, so if they could profit by denying coverage of drugs that are less effective they might pay for it.
This could also have been a good charity. Providing helpful information is what educational charities do. Making a nonprofit would save on costs, and might have been enough to make this viable (though certainly not a lucrative money-maker).
As you're now 18 months on from July 2018, so have probably moved onto other things...
Please consider approaching some academic institutions, or some medical professional associations to see if they would be interested.
While I doubt any will have any money to pay you, they will almost certainly be able to host this for free, probably indefinitely. You just need to find one interested researcher.
It's hard to tell exactly what it is you made, but informative, interactive, visualisation tools can have really powerful public health effects if marketed well.
I'm surprised Big Pharma hasn't done something like this for free for doctors. They already spend tons of money marketing to doctors to perscribe their drugs and even club together and spend billions monitoring the perscription data with IQVIA (IMS Health basically). Maybe it's just that I've recently finished reading Bad Pharma by Ben Goldacre, but this sounds like it's right up their street to help manipulate doctor's perscribing.
Hey, hey -- what about a grant from WHO, NIH or NHS? What about having a national or world government continuously sponsor this work as a public resource?
NIH grants are extremely hard to write and get funded if outside of an educational institution. I have tried both in and out of school; inside it was relatively easy, outside near impossible. YMMV.
This is an amazing journey. You got a novel idea. You felt the fire under your belly to bring it to life. You executed that idea and really built the product that people like. So sales and marketing are hard. So what. You have learnt a lot more. The money was not wasted. It's part of a very valuable education and on-hand experience. The lessons learnt will benefit you time and time again.
> There was this giant thing called healthcare right, and its main purpose is improving health—trillions of dollars are spent trying to do this.
There was this giant thing called healthcare, and its main purpose is applying well known treatments to people whose symptoms more or less indicate that treatment might somewhat improve their condition—trillions of dollars are earned trying to do this.
For anyone else who, like me, is prone to the excitement of new company ideas and has a tendency to gloss over questions like "is it even possible to monetize this?" I'd recommend reading Disciplined Entrepreneurship. It's not the One True Way but the steps they recommend cover a lot of questions that you should think about during the process.
Here's how I see it, and the quote at the end kinda fits into this:
What problem are you trying to solve? Googling "what is the best pain killer" is a very "average user" question. It's also 99% useless.
Pain killer for what? Headache? Muscular pain? Nerve pain? You have a partially severed limb that's bleeding profusely? Oh you mean an OTC pain killer?
Ok do you have stomach problems? Liver problems? Kidney problems? Diabetes? "Oh I don't know" oh well...
"But I'm just trying to get the best drug for your situation" yeah, that doesn't exist. And there are personal preferences (doctor), personal preferences (patient), limitations, marketing pushes, etc.
Finding out drug X is more efficient than drug Y in 5% is just background noise. Hence why the doctor didn't see much value in this.
What would be valuable is finding which drug X to try if drug Y didn't work or can't be used (though doctors usually have that in their heads).
His problem is that this product neither increases the revenue of the practice nor does it reduce(or remove) any existing cost. Without that it's a cost center without clear benefits.
> It had been a bit of a working assumption of mine over the past few weeks that if you could improve the health of the patients then, you know, the doctors or the hospitals or whatever would pay for that.
No. Or at least you better show clear benefits becuase EVERY medical product makes claims that they improve patient health or patient outcomes.
What he needs to do is figure if there is some costs that are currently incurred by hospitals and practices that could be reduced or removed by using his product. If he can do that, he'll get sales (and in fact, there is a lot of money in 'population health' and related fields). Otherwise, he's got nothing.
Thank you for sharing the details of your experience. Healthcare in the US is not a normal market, so normal product- based thinking doesn't work. The end user doesn't pay, so they cannot directly influence the market by choosing products. Your discovery with the doctor who said that improving care doesn't help her, is really insightful. The 80/20 insurance rule fixes insurance profits so they have an incentive to make care as expensive as possible. Despite all this, we all know there is a ton of value in making things better. Your graph of Aleve performing so much better then ibuprofen is really interesting, and seems to capture how other parts of the system could be improved if end users were given the choice. Maybe a trusted review site with data to help patients choose doctors.
Hey, I prototyped something similar to this a while back.
It used NLP and Elsavier search API to do entity detection and sentiment analysis on drug/effect pairs.
So you would search like "Mirtazapine", and then it would create a word-graph where the drug was in the middle, and it would be connected by things like:
The words were colored either green/red based on whether it was considered a positive or negative effect, and the size was based on the size of the effect and numbers of mentions.
It was meant to be a tool for clinicians and prescribers to be able to enter a medication and get an easy overview of its effect.
Do you have any interest in pursuing this further?
I know a few investors who I think might still be interested in this.
This would be a useful feature of Google, but it won't be a real company without an innovative business model. 9 months to build a company AND THEN innovate a business model is nowhere near enough time, especially since you probably spent the vast majority of the time on the product. If you had a partner that was equally talented working on the business side since day one you still wouldn't have something in 9 months, but maybe 2 years. And maybe something HUGE. That's the gamble you make going into a startup, and 9 months simply isn't long enough to find that out.
Or maybe you could have been a small acquisition target for Google to incrementally increase their search superiority. I imagine they would be able to monetize eyeballs with searches like these at (very) roughly 100x the rate of WebMD.
I have had life long headaches that I have given up on fixing. Occasionally when I switch to a new doctor I will mention it just to see if they have anything new to say.
They usually just put me on a concoction of various pain killers (e.g., migraine specific stuff like sumatriptan). These prescriptions actually work great. But so did OTC aleve/advil/tylenol too.
The problem was that I didn't really care for more effective pain killers. What I really want is to fix the root cause and stop having these headaches.
So yea, I can totally see why something like this would fail. If the solution is still qualitatively the same, a small percentage improvement is just micro-optimizations that most doctors and patients won't care much for.
It seems like a flaw in the system though because those micro-optimizations are what usually drives the free market and accumulate to become huge. If all buying decisions were just made on the same coarse decision making then the useful competition-by-endless-small-improvements would be replaced with competition-by-insidious-marketing, which I guess is the state that pain medicine is in.
Pretty spot-on guess! I do have relatively high blood pressure, typically in the 130-150 range.
Doctors have suggested it before but since I am borderline I don't really need to be on meds yet. I think it sounds worse than needing to take pain killers occasionally.
I was borderline. Higher bp increases your risk of stroke. I figured the headaches were a warning.. I could be wrong about that though.. I take a very low dose ace inhibitor and it works great. Doc said it wasn't a permanent thing necessarily.. I could probably go off of it if I was more consistent with the DASH diet and exercise more and drop 15-20 pounds.
Keto diet supposedly helps drop the bp too but diets are really hard to maintain and stick with.
Healthcare startups are a like a 'land war in Asia'... "only a fool would make such a blunder" (via The Princess Bride, perhaps the best film of the 1980's.)
Selling anything to medical practices is remarkably difficult, due to the difficulty of getting a meeting. They may have to lose $250 in billable time to meet with you, and they are bombarded with sales calls from everybody under the sun. Getting adoption is worse, with overworked staff actively hostile to change. I watched a friend sink his heart and cash into a medical research survey system only to find out how impossible of a market it is to break into.
And it is a land war. Selling into large companies is hard fought, but possible. The long tail is practically impenetrable for a software system, not when they could allocate capital to revenue generating marketing, equipment, and staff. I worked at trying to market a low priced medical supply item into a long tail of allied healthcare professionals and found that there is virtually no network effect. That is to say, people who share the same medical profession have very little interconnected outside of a very local level, and the work they perform is so physically and mentally draining that they don't really want to spend a lot of time online after hours talking about vendors, or talking with others in their field at all. Durable medical equipment and consumables sales are really driven by incumbent field sales reps more than ecommerce, even still in 2020. Land war.
Look at how difficult a time big tech has had taking on electronic medical records systems. Technology is not the bottleneck.
There are plenty of health tech startups finding traction, and I do believe medical professionals want to continually improve themselves and their practices with whatever tools and information is available to them. The fact is, $40K isn't enough capital to win a land war, no matter how great the solution is (and its looks great). I'm sorry OP had to find that out the hard way.
PS. There is a growing array of health focused startup accelerators that can provide seed money, but more importantly the expertise to navigate the sales and regulatory challenges unique to healthcare.
The foremost problem is lack of domain expertise as others have noted. If you lack domain knowledge but love your idea, the best course of action is to keep your day job and gain as much knowledge as you can before quitting.
But another problem is lacking the conviction and traversing the idea maze[0]. Several comments here have provided good paths for possible methods of execution like offering a free API and seeing who is using it.
Why GlacierMD? You know glaciers tend to melt these days.
You'd be surprised that the dumb and not so noble idea of your buddy with Doppelganger could gain much more traction and make much much more money than your idealistic thing... this is usually the thing.
People are attracted more to fun than to pragmatic usefulness.
You feel almost like a martyr who wanted to save the world, but nobody understood you. I am very sorry and I see great value in your idea, but looks like people don't want to consume that.
Indeed, why would a medical practice want to completely cure the patient and never see them again?
Really fun read and with lots of useful references.
Oh man this is depressing. I have a few ideas that I'd like to get funded and make a business out of, but I have no idea how useful they would be to the average consumer and how I would go about raising funds (doesn't help that my current job has a pretty strict NDA which pretty much covers all software).
At least you gave it a go; even if it didn't work out, you don't have to wonder "what if?" for your entire life, and I think it's pretty cool that you built something you were passionate about; at the very least, stuff like this is pretty good resume fuel, right?
> It had been a bit of a working assumption of mine over the past few weeks that if you could improve the health of the patients then, you know, the doctors or the hospitals or whatever would pay for that.
Wow. That's not how any of (especially 'for profit') healthcare works.
I think it's a good heuristic that you definitely shouldn't build a product if you rely on getting money from 'some profession or some institutions or whatever'. This indicates that much more research into your future customers and motivations is needed before you attempt to create business.
HN with the editorial decisions on the title again. What's with this?? Seems good enough to go with the author's actual intended title for a few hours on the HN front page, and now it's changed?
I love the idea, but I'm sad that the author focused on doctors, and not ordinary people that he mentioned at the very beginning. This could be a great application used by everyone, including Google, to provide better search results, at the same time providing more and more traffic. Mentioned ad revenue could also work, but not only this.
If the author reads this, do not be like Google, don't give up, just pivot. Yes, I think it's Brilliant Startup Idea that could be back in a few months better than last time. :)
'“To succeed, an offering must create value for all entities involved in the exchange—target customers, the company, and its collaborators.”' aka product-market fit
You could have made value for the customer too, you needed cash to keep the operations going. Doctors will pay for this if it lets them get a leg up on other doctors that's how you need to sell it. "Doctor X prescribes what he always prescribes, I make decision based on cutting edge studies." If they got arbitrage because of that and could steal patients, especially doctors with a new practice then you have a sale my friend. You gave up too easy.
One of my good friends is a real estate agent and he is helping me find a house. My job is to complain about every place we see. Every single time he finds a way to spin my complaint into a positive or something constructive. Roof leaky? You can get it fixed for $X dollars and we can work that into an agreement.
For a great salesman there is no objection that cannot be overcome.
A lack of money from the customer can’t be overcome. In your case if you can’t afford any of the houses there is nothing the great salesman can do. Of course a great salesman qualifies their prospects before attempting to sell.
In the specific case of the article, the developer was reacting to the doctor saying: "So I’d sorta just be, like, donating this money if I paid you for this thing, right?" This was not a price objection type statement.
Also, selling houses to people who did not have enough money for a house was the basis of an entire financial crisis. It looks like creative salesmen found a way around that.
You can argue some extremes if you want but exceptional salesmen are expected to sell ice to eskimos and sand to arabs. And I would expect them to handle the pretty simple objections raised in this article, especially those around price concerns.
Yep, same here. I've always just used whichever OTC pain killer was most convenient and have never noticed a difference between them. They all seem mildly effective at reducing discomfort. I have taken Ibuprofen for headaches many times and it seems just as helpful as anything else.
Others have said this, but $40K is a bargain to get this experience. If you include the opportunity cost, my own 4 year start up failure cost me way more than this (even if you don't include lost opportunities it was probably about this sum of money). You still have the database/website that you can continue to run in your spare time and you never know it might even make you a return in future.
I was given some advice long ago that I stick to, and which avoids situations like this entirely -- if a startup cannot point to a repeatable transaction where they get revenue in exchange for something... then it not a business, it is a side project.
Admittedly, side projects can grow into something over time, as you evolve the idea. But there is zero reason to quit your job for such ideas.
So, given the author quit his money making startup to make a business that would be helpful for humanity, I assume that all of his data and IP is now in a .zip file up for auction and not actually disseminated to people who could use it right? It tells you that what people say is their motivations, especially when it comes to startups are often not at all the case.
I'd go further and say, if this could provably generate ad based revenue from a non-trivial number of users, Google would build the same thing and include it in the basic search result shown in the article (with adwords suggestions) for effectively zero cost to Google.
So, even if it were financially successful, it wouldn't likely survive the competition.
"Doppelganger has created value for the customer but not for the company."
Isn't the value for the company created through the value for the customers? I.E. you take a percentage of the value for the customers as the value for the company.
What would be an example of creating value for the company?
I'm not convinced the idea could never possibly turn into a successful business.
Maybe he just tried to sell to the wrong people, maybe pharmaceutical companies could have been interested, maybe re-packaging the product differently...who knows...
The fact that it could provide real value for many people already sets it apart from most ideas.
Company that made Doppelganger-like product is flourishing. They are selling it`s product to governments and other companies for ton of money.
https://en.wikipedia.org/wiki/FindFace
So many mistakes here in both concept and execution. That conversation with the Psychiatrist is bizarre, and must have been embellished for a good story. I think what 'Susan' is actually trying to say is that the incremental benefit of using this 'auto meta-analysed' data over the actual studies is very small. The author tries to cast this as 'Doctors not wanting to improve healthcare', which is just a rationalisation for the product which is not useful. Case in point, the spreadsheet cited at the start of the article largely links to a systematic review that is already published, ie the author is copying data someone else has already aggregated and then re-analysing it in basically the same way with a few more studies. And then trying to charge people to access it. This seems manifestly inferior to the existing system if you ask me. Cochrane would be turning in his grave.
This could have worked. It would have been possible to sell an integration with a major health insurance company or provider network (or a few) as a tool available to their members.
Like as a United Healthcare sponsored portal that helps you pick OTC drugs, or similar. Those guys love stuff like this. And I've actually worked in this space so I'm not just making things up.
With that said, the way to do that would be to make it a destination with a growing userbase and then approach them for partnership. Which would have cost money and taken a while to ramp up, so you'd need some runway. Also it's not a strategy to actually get rich, that wouldn't lead to tens of millions in revenue or valuation probably.
But it was in fact a good idea and there were some paths forward you could have taken. Had you done so there's an excellent chance you would have discovered some interesting and perhaps lucrative product extensions along the way.
If the technique is solid, they should sell to big insurance companies. That are the actors that are interested in paying for the most efficient treatments. At least when it comes to choosing between short and efficient or long lasting.
Very nice write up! I can easily see my own biases leading me down some of the paths you mention, so thanks for sharing. A more optimistic way to look at this situation is that you spent $40k to get to the front page of hacker news!
Agreed, this was delightful to read! I really appreciated how the author was open about naivety and blind spots. It made the contrast more apparent when they learned things. I'd love to read more by this person.
This story is fantastic, honest and educational. Clearly the author is smart, yet this doesn't prevent him from making some stupid mistakes. I know it is obvious, but this story is very educational.
Failed entrepreneur, perhaps, but one hell of an entertaining writer and storyteller while throwing in bits of wisdoms. I also came away with one insight at the end. I'm going to sign up for the news letter.
Could you try to sell it to insurance companies? They might find value in providing it for free to clinicians because they might save money with better outcomes / preventing follow up visits.
Insurance companies could also use the data to argue with healthcare providers about what the best cost vs value is on a course of treatment, but they probably already have data scientists doing this kind of work for them. Insurance adjusters and all that...
Insurance companies don't care about outcomes - most are structured in a cost plus model where they make the "plus" - so if anything, something that costs more is better for them.
this is quickly changing. For insurers that offer Medicare Advantage plans, they get paid on a per-member per-month model, so outcomes are very germane to their profit margin.
he "worked at a startup that leveraged autonomous blockchains to transfer money from naïve investors to slightly less naïve twenty-somethings. There are worse gigs."
To get in the door with doctors you need to pay them to use your product and then sell the data they generate to other companies. (not HIPPA data, but metrics, etc)
did you try selling to pharmaceutical companies what about healthcare provider networks what about EMR companies? they would probably compare you to open source tech R packages, but it's worth a shot.
you could make a "reference database" out of it and charge for access to the latest version
This was my reaction ... it's presented in the article as a fun idea without a market, but seems like a plausible play to build a saleable corpus of facial recognition data.
It's an excellent write up and there's great lessons in there for all of us. And, it's great that he started talking to actual doctors, if only he had done that first. I know i've made that mistake many times.
As for the medical industry: the idea is to keep people sick as long as possible. You can't make money from people who get better or people who die.
At first I thought this was satire. Here's this guy, who uses a pretty straightforward analysis without any prior healthcare domain experience, thinks he's found the holy grail after a few weeks of work, and then spends all of his money (without first talking to actual domain experts) before he gives up, because Healthcare.
"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."
Thoughtful critique is good, but this post breaks the guidelines and falls outside the kind of thing we're going for here. I know it's the internet default generally, but please be more thoughtful when posting comments to HN, especially about other people's work.
This is why I get so grumpy about startup narratives. If you ask successful unicorns how they got started, they'll usually explain that's exactly what they did. Garrett Camp was just trying to hire a private driver, and realized it was too hard, there's no reason it should be so hard. Jack Dorsey saw a guy who didn't accept credit cards and realized it would be great if he could accept credit cards.
If you don't realize what's going on, that companies pick their founding myth based on which story sells the best rather than which story occurred on the earliest calendar date, it's very easy to get a false picture of how good startup ideas originate.
This. Say what you will about Bezos, but he seems to be the most impressive tech founder out of any of FANG and any of the large unicorns founded after Facebook.
Such an impressive list of deft moves that he made (in a range of different areas) one after another between '94 and 2000 to set Amazon on a path to meteoric growth.
There's a lot of survivorship bias there. Case and point: the author did have this specific problem, and surely this problem is common.
Uber 'bought' their market - it might not even be viable when they run out of dumping money and have to raise prices.
I think there might be some opportunity in this guys startup with some pivoting.
Also I think there was some funniness on the feedback: Doctor used 'cost' as an excuse, but it may have simply been that the product was not hugely valuable.
Maybe for common drugs it won't help, but there are a zillion drugs out there, maybe doctors want a good reference.
The revenue model he didn't articulate comes from the drug companies. If Doctors used his tool, drug companies might pay for it in some way.
Finally, I have to say 200K lines of code sounds like a lot ...
Part of being a successful founder is spinning a yarn for your startup though. It's not improbable that the majority of successful founders are also just good at telling the story about it while glossing over the boring hard parts.
Couldn't the founder story still be true, but simply incomplete? E.g. Jack Dorsey may have been frustrated that a company didn't accept credit cards, but also, he was already a successful entrepreneur and may have spent months or years researching the problem before starting Square?
A lot of these founder stories boil down to coming up with a good idea and then building it. You hear more about the stories where that worked out and less about the stories where it didn't. Plus the ideas and execution quality all vary, along with market timing.
It's hard to say what exactly can be learned from such stories.
I think the _"thinking he's found the holy grail"_ and _"eating Domino's at his parents basement"_ were added more for flavor than anything.
He had an idea, that seemed good, and invested $40k that apparently he could afford to loose. He made some mistakes that could be avoided in hindsight (like everyone trying something new).
The dreams of being billionaire and stuff is just to make the reading fun (it did for me).
Obviously, the doctor is going to torn to shreds here...but I think this kind of misses something about medicine. You prescribe a drug, you know roughly what it does...but aren't totally sure what is going to happen when the specific patient takes it. SSRIs are a perfect example of this.
...and that is kind of why you hire a doctor rather than a computer who has looked at some studies...that isn't really what medicine is.
More generally: this just sounded like a terrible business model. Consumers and probably doctors would value the end product but are you going to get back the money spent on contractors manually going through these studies...no. I can see this making sense for a larger company but these kind of projects are terrible for one guy alone, they just tear through cash.
You also don't seem to have asked how doctors/insurance companies actually look at this data either. I am sure I have no idea but the first question should be: is the user looking through these studies manually? Did you survey doctors: how many studies do you look at a week? Do you look at them before prescribing? Etc...and probably do this before you build obv (and try to get money upfront).
Falling in love with an idea plus ignoring reality is the surest road to hubris and failure. Be religious about results, customer value, profitability and employee welfare, not ideas. Ideas are a dime a dozen; executing and pivoting into what works too often is stymied by ego and thousands of other self-destructive tendencies that increase the likelihood failure in millions of ways. There are very few ways to succeed, with a little help from happy accidents and flexible resilience under pressure.
If you are interested in creating a startup, you should first find something that can actually make money and then work on creating a great product around it.
Too many people come up with a great idea and then hope for the best when it comes to market fit and actually getting customers and achieving profitability.
It was really good to read some self reflection and analysis on why you failed. Most people ignore the actual reasons their ideas/businesses fail and instead blame it on someone, or something... Most of the time, its the fact that there was no real market for their product!
Silicon Valley doesnt necessarily suffer from a lack of engineering talent as so much a lack of marketing talent. Far too many ideas go to market without any understanding of the actual market. But hey, people get to say they were a CEO for a few months and that sounds great on their Tinder, I mean LinkedIn profile. Right?
Unfortunately, I feel like this is a lesson that's best learned the hard way. Everyone can tell you from the get-go to "have a business plan" before committing financially and professionally to a self-concocted venture. It's not until you run through it and witness the brutality of failure first-hand that you realize the importance of doing your homework first and having a plan.
And here's something that's absolutely critical for all engineers trying to build a startup in healthcare to really understand. Healthcare is so alluring because tech people are so idealistic. We think, "Wow, we we can write some code, and then save lives!" And that's sometimes true. But when you are thinking of your customer, revenue model, sales strategy - all that business stuff - the mistake we make is thinking healthcare is different and the same rules don't apply. We think if we can save lives, we can make a business. But the key insight is that you have to take a step back and just treat healthcare like any other for-profit business. Doctors and hospitals care about what generates revenue. Insurance companies want to save money. Pharma companies are looking to advertise to new customers. It's no different than other industries. The psychiatrist's reaction to the sales pitch is the classic thing a doctor will tell when you're not helping their bottom line. As I was reading the post, I could almost predict how she was going to respond.
And if you figure that business stuff out, who knows, you might just build a healthcare business that saves a life or two.