I spent $ 40 thousand and ruined a great idea for a startup

Business doesn't always come out of a good idea

The head literally splits. You are standing in a pharmacy, massaging whiskey and looking for something on the window – anything – just to stop this hellish pain.

What to take? Tylenol? Advil? Naproxen?

I suppose most will grab something cheaper, or whatever comes to hand, or something familiar. But if you're scrupulous enough, ask Google for the best over-the-counter pain medication (OTC). Here's what our friendly tech hippo will answer:

Oh thanks google you listed all options.

If you google questions about health, like 77% of Americans, such a vague answer will not surprise you. Although in fact he should to surprise, because every year researchers conduct tens of thousands of clinical trials. And hundreds of clinical trials are devoted to the effectiveness of pain medications. So why can't I google these results?

And in blessed 2017, I was born Startup Brilliant Idea: create a structured base of clinical trials to extract simple practical answers to common medical questions.

As a proof of concept, I tried to do it manually: I compiled a spreadsheet with the results of all the tests of over-the-counter painkillers I could find. In R, he wrote a program for network meta-analysis, the gold standard for evidence-based medicine.

The result turned out to be quite interesting – exactly what I was looking for, aimlessly wandering my eyes in the pharmacy window:


The most effective painkillers. Results of the first meta-analysis (157 studies, 4,400 people)

A wave of excitement swept over me. Here is the problem that

  1. Interesting
  2. Can help people
  3. I know how to solve it

The perfect combination. After hours of searching for free domains, I came up with a name for the project: Glaciermd.

Over the next nine months, I will quit my job, write over 200,000 lines of code, hire five contractors, register a legal entity in Delaware, invite four doctors to my advisory board, and demonstrate GlacierMD in twelve Silicon Valley healthcare facilities. I will spend $ 40 thousand from my own savings on the purchase of clinical trial results and the remuneration of contractors to add these results to the GlacierMD database.

On July 2, 2018, the GlacierMD engine performed the world's largest meta-analysis for the treatment of depression, using data from 846 studies, breaking the previous Cipriani record in 522.


GlacierMD Depression Fact Sheet

A choir of angels sang in my head. Here I am, embodying Silicon Valley's dream: improving the world with technology.

Two weeks later, GlacierMD closed.


“This is an amazing idea,” Carl said. “It seems worth working on it.”

Carl is my boss. We worked in a startup that used autonomous blockchains to transfer money from naive investors to a little less naive twenty-something people. Business is worse.

And so Karl says that the idea of ​​a startup will bring such benefits to humanity that I’m just should leave, and let his deadlines burn. I nodded in understanding, feeling how the weight of responsibility fell on my proud shoulders.

“Thank you, Carl,” I said, “I will try to mention you at the Nobel Prize ceremony.”

Two weeks later, I quit and began writing code at a frantic speed. He drew with a marker all kinds of incomprehensible diagrams on the windows of the parental home. Hired a motley team of Egyptian contractors to start entering clinical trial results into a database. I ordered a logo, registered a domain and plunged into color schemes.


Data Extraction Portal for GlacierMD

When I finally made the first alpha version, I showed it to the product manager at my former company. I watched him study the demo, and waited for his eyes to expand with surprise. Instead, he simply shrugged.

“There are a lot of medical advice on the Internet,” he said. “Why should I trust yours?”

I started chatting about network meta-analysis, statistical significance and p-values, but it interrupted me:

– Yes, everything is great, but nobody cares about this mathematical shit. You need doctors.

Heck, he was right. If no one cares about math, then I am no better than Gwyneth Paltrow with her eggs. To earn trust, you need to get recommendations from reputable people.

So I called some of my friends, pals, friends of my friends: “Would you like to become a consultant for my advanced startup in the field of medical technology?”, I asked, eating pizza in the basement of my parents' house. I can offer 1% of an extremely valuable, fast-growing startup, in exchange for your faces throughout the site.

Four doctors agreed. It is called to make dealsladies and gentlemen, I felt like a favorite pet of Warren Buffett and Dr. Oz.

Everything is going fine. Friends and relatives unanimously say that they like the site. Even some strangers on the Internet liked him.

“I know that this is truly truthful information,” I tell them. “And how much are you willing to pay for it?”

However, the question for some reason causes the same reaction:

– Ha ha ha ha! Well you said!

I forgot that the first law of consumer technology is no one pays for consumer technology. But the problem is, I tell myself. That's why Eric Schmidt invented online advertising. I’ll just put some banners on GlacierMD – and bim bum boom, I'm already in the floating city with Peter Thiel in front of the Burning Man.

But then I looked into the financial statements of WebMD and was sad. It turns out the world's largest health site earns about $ 0.50 a year per user. That is … not enough to launch GlacierMD. I spend on rent, on Egyptian contractors, on AWS – I will soon need some kind of income.

What I need is users who are willing to pay for this service. What about the doctors? Doctors have money, right? (In the US, doctors have a high income – approx. per.) Doctors, trainees, or maybe someone else from the medical industry – maybe they will fork out for my custom meta-analyzes?

So I listened to a few podcasts and became a sales expert. I started a cold scripting on the Internet, trying to convince people to watch the GlacierMD demo.

In the meantime, several troubling messages came from Egyptian contractors.

“I think it's time to talk about raising salaries,” one wrote.

“I feel that I have become exceptional in my work,” wrote another. “Please consider raising salaries, or I will stop working.”

“Please increase the payment,” the third one directly wrote, sending useful screenshots with a demonstration of how to make this increase through the Upwork website.

Are my contractors unionized? Amazing I catch a glimpse of the current account balance that is leaving at the peak, gritting my teeth and approving the increase. At this rate, money is enough for just a few weeks.

But my sales calls are starting to pay off. Miraculously, several doctors agree to speak with me. So I take my parents' car and drive to the suburbs to meet a doctor, which we will call Susan.

Susan has a little practice in the center of Redwood City, a town in Silicon Valley, as if printed out from pictures from a search main street.

Susan is a little chatty (she's a psychiatrist), but in the end I still run the demo. I show how to filter studies according to the demographic data of the patient, how to get treatment recommendations based on the profile of side effects, how to create a dose-response curve. She whining and gasps in all the right places. By the end of the conversation, Susan was practically drooling.


Comparison of the effectiveness of two antidepressants over time

Hook, fishing line and sinker, I think to myself. I'm already wondering which Away bags will look better in the back seat of my Cybertruck, as Susan interrupts my train of thought.

– What an interesting project! She says enthusiastically.

Something in her tone makes me stop: “Well, yes,” I say. – And imagine a similar product that can change itself practice medicine – how much would you pay for such a service? ”

“Ahhhh …” she said. – Honestly, I don’t know if we can spend the budget on this. The program is very interesting … but I'm not sure that our practice will pay back these costs.

From sales books, I know that if people talk about too high a price, then in fact they mean that your product not valuable enough. But before that, Susan was so high as if I were offering her Nirvana-like-service, so the conversation took a rather strange turn.

“So you don't find this product useful?”

– No, of course it is useful! I mean, in many cases, I just prescribe what is usual, because it is convenient for me. But maybe sometimes I will prescribe something else based on your meta-studies.

“And does it cost anything?” Know the best treatments?

“Hmmm,” she thought, picking her nails. – Not certainly in that way. Of course, I always care about the best interests of my patients, but, you know, they will not pay more if I prescribe Lexapro instead of Zoloft. They will not come or bring more friends more often. That is, I kind of throwing away money, if I buy this thing from you, right?

I literally had nothing to answer. Over the past few weeks, I have developed some working hypothesis that if you can improve the health of patients, then, you know, doctors or hospitals, or someone else will pay you for it. There is such a giant thing called healthcare, and its main goal seems to be – health improvement. Trillions of dollars are being spent on this. So if you created a program that improves health, someone has to pay you, right?

I said goodbye to Susan and tried to perk up. There were still a dozen meetings with doctors all over the bay – well, not all the same ruthless capitalists like Susan. Maybe they will see the brilliant superiority of GlacierMD and lay out some money.

In fact, everyone answered in the same spirit:

“We simply will not justify the costs,” the pediatrician admitted.

“Not sure if this is included in the budget,” the therapist replied.

“Terrific,” one of the hospital staff said. – You should try to sell it! – ugh …


Therefore, in July 2018, after nine months and $ 40 thousand spent, I closed GlacierMD. Dismissed contractors, archived the database and disconnected the servers. GlacierMD is dead.

Do what people want. This is the motto of startup incubators and the main maxim of novice Internet entrepreneurs. The idea is that if you create something really amazing, you will find a way to capitalize on it anyway.

Therefore, I did what people really need. Both to patients and doctors, yes to myself need such a service. Where is the mistake?

Sometimes I like to disconnect from the Internet pseudo-dropper – and read books. I know this is strange. Maybe a little hipster. But recently, I read Wharton’s beginner’s marketing tutorial, Strategic Marketing Management. This is stated in the very first chapter:

To succeed, the proposal must create new value for all entities involved in the exchange – target customers, the company itself and its employees. – "Strategic Marketing Management"

For all subjects. You cannot just create value for the user: it is charity. You also cannot simply create value for your company: this is fraud. Your goal is to create a kind of exchange with a positive amount, where everyone wins, including you. According to this tutorial, a business plan begins with a simple question: how do you create value for yourself and For the company?


What is taught in a business school

I involuntarily flinched when I read it. How much time could I save! If I formulated the GlacierMD revenue model from the very beginning, I could research the economics of the advertising market or verify that doctors, hospitals or insurance companies are willing to pay for such a service.

A few months after the closure of GlacierMD and return to corporate life, a friend offered me the idea of ​​a new startup:

“Let's call him Doppelganger,” he said. – It's very simple: you upload a selfie to the database, and he uses artificial intelligence or something else to instantly find in the database all who …

“Looks like you,” I finished for him.

“Exactly,” the friend nodded, grinning with all his teeth. – Awesome! We have to do it!

Well yes, that sounds like an interesting party joke. In the narrow sense, I would like such a service, but I am not going to devote any time to it for anything. Doppelganger creates value for the client, but not for the company.

“Call me when you write a business plan,” I said, laced up the old crosses and drove off into the sunset on a Lime scooter.

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *