Once upon a time, back in 2016, Welltory started out selling a lifestyle tuning service to rich biohackers for $99 a month. For people who found it easy to decide to buy a pile of gadgets, run any test you asked them to, and who were already willing to invest in their own health. They collected data with our app, we built a ton of integrations, and they talked to us through an Intercom chat.
We ran that service for a whole year. The press wrote that we’d created a profession from the future, the personal health analyst. That’s also where we found out that nobody teaches healthy living anywhere, doctors included, and practical analysis of health data least of all. So we opened training courses for personal health analysts, with a job at the end of them.
We got our first traction. That’s how we earned our first few hundred thousand dollars. We tried really hard. We made super complex, super detailed reports for our users. We made them programmes, training plans tuned to their individual quirks, all of it. We built enormous knowledge bases. But most of all, we talked one-on-one with our users, with two specific goals: see a measurable result, and get the subscription renewed. If you can’t show the result, you don’t get the renewal. Very simple.
What a year of talking to people taught us
Talking to people directly and delivering the service yourself is like non-stop customer development in a consumer market. That’s how we learned that nobody actually wakes up in the morning thinking “what could I do today to lower my risk of a heart attack in ten years.” We understood how much the problem sits in the head (knowledge and beliefs) rather than in the fact that someone had no training plan. We saw how much the results differed between analysts, where it was obvious that their personal “influence” and powers of persuasion decided more than the factual, theoretical part. Obviously I’m not talking about the cases where we simply found that something was wrong with someone and sent them to a doctor in time; obviously in those moments you earn a different kind of loyalty. But in general it was clear that the individual factor matters a lot, and that this was going to be very hard to scale. We really were helping people. They paid us gladly. We tried really hard. But we had no idea how to scale this 100 or 1000 times over without losing that quality.
At the time ginger.io didn’t even exist yet, nor did that whole pile of health coaching services. They all showed up after we’d already decided this was a dead end from the point of view of the company’s mission, which is to help a large number of people improve their health. That’s how we understood we had to go software only, so it would be cheap enough to help many. But without that year we’d have started building the app on completely unrealistic assumptions, the ones that killed an enormous number of wellness apps. That if you give a person a personalised plan, they’ll do something with it. Yes, you can make money that way, on hope, on a promise and on a person’s guilt (right, I’ll buy the plan now and theeeen I’ll start doing something), but you can’t actually affect the health of a large number of people.
That’s how our path as an app began. All that’s left from our history is Intercom, embedded straight into the app. It’s actually a B2B support service, consumer apps almost never use it, and we use it at 2000 per cent of its functionality and have kept a live support team this entire time as a matter of principle, one that answers and talks to users far more closely and far more knowledgeably than usually happens. It got harder as we grew, but at some point AI came to the rescue, and the combination of well-prepared, motivated people plus AI gives us that live feeling of being talked to, which apps that don’t invest in it so often lack.
So here’s what I’m getting at
On our ten-year turn we’ve come back round to the beginning. We’re opening a tier with direct, personal, live care again. With the founder and our scientists directly involved in shaping the knowledge, the hacks and the innovations. With users directly influencing the product. With people who actually know this material, on hand all the time, instead of a fifteen-minute call with the cheapest available doctor, the way it works in telehealth.
It’s called Care, and the first group is for people with energy limited conditions.
As we grew we gained something almost nobody else has, thanks to a large base of users with connected data. At eighteen million installs there are finally enough people with data in one place that we can find similar people not just in terms of one diagnosis, but in terms of physiological pattern, similarity of lifestyle, and comorbid state (the combination of diagnoses). That’s not something you can do at a few thousand users, and it’s the one thing ten years bought us that we couldn’t have bought any other way. We use it to form groups of people who are actually like each other, and try to help them as a group, with group cases and group access to experts on their conditions, to analytics on their data, and to each other’s experience and knowledge.
Users influencing the product is the part that’s easy to say and expensive to mean. What it means here is that a question raised in a group becomes engineering work, and that someone’s condition and constraints have to change which tools they get and how their data is read. Selling a community alongside an app that stays the same would be a much easier business, and a dishonest one.
For me personally, the most recent piece of work I’m proud of was a report for doctors for people with energy limited conditions, which I learned to make while solving an unsolvable problem: how do you get something out of wearable data that a doctor will find both legible and useful? They aren’t trained to look at wearable data, and very often simple numbers like variability are actually useless on their own. And when our users started bringing me reactions along the lines of “huh, now that’s interesting,” I understood we were getting somewhere.
I mean, any fool can make you a stupid report for a doctor. You can go print one out of any wearable app. Making one where the doctor says that’s useful, that’s interesting, that’s the hard part.
The group format also solves something very important: it moves the emotional influence that used to depend on that one personal health analyst onto a group of people who support and understand you. That makes it potentially far more scalable. What you get is a kind of patient-led cooperative care, supervised by medical advisors, scientists and analysts.
Who we’re doing this for
Our target audience now is people whose overall quality of life depends very heavily on the fine details of how they live. Energy limited conditions, people with migraine, people with irritable bowel syndrome, people with panic attacks.
In reality science knows that what happens to them depends on fine management of stress and of intellectual and physical load, but that management is made difficult by the fact that we’re all different, every combination of conditions creates its own requirements, different regulatory loops are damaged in different people, and in the modern world there simply are no mechanisms and no universal pills that will tell you what to do specifically for you, taking into account everything about your particular life and body. You stand in front of this challenge alone, with no preparation, no knowledge, no expertise and no help, and very often in complete solitude and without support, making your peace with the fact that from time to time you lose days of your life, spending them in bed, in a dark room or in the bathroom because you simply can’t leave it. Or you simply can’t get up. Or you simply can’t even think, because your head is splitting and the pain is pulling the soul out of you.
One of our users put the shape of it better than I can: to get to the doctor tomorrow, I have to take my shower today. Two ordinary tasks in one day is too much load. That isn’t a motivation problem or an information problem, it’s budgeting, against a budget most people don’t know they have, and no general model of stress and recovery has any concept of it.
And we, as people who’ve been studying habits and physiological data for ten years, have a lot more ways of helping them. More than their doctors, or clinics, or research groups, which often can’t even recruit 200 people for a study. And the main thing is that phenotypic groups this small aren’t interesting to big investment. They’re too narrow for anyone to bother building for them as a group, and too varied for universal pills to work on them. Which is why the most popular software in the pain market is just apps for keeping pain diaries, in an attempt to somehow help people find some correlations.
And that’s nonsense that doesn’t work. Because regulation doesn’t work that way. While people with migraine are busy working out that the attack happened after one extra cup of coffee, they aren’t thinking about the fact that there was actually a two-week downward trend behind it, where that cup of coffee was just the last straw, and the last straw could have been anything. It’s worse than that, even: the prodrome, the premonitory phase, starts hours and sometimes up to two days before the pain, which means the chocolate or the coffee you blamed was very often not the cause but a symptom of an attack that had already started. And millions of people are tracking their migraines with these apps, blaming themselves for not being able to find their “triggers,” when it’s simply the wrong question (and scientists, by the way, know this).
Recently my CMO wrote to me that she’d gone a month without an attack for the first time in her life. Because she watched the metrics in our app closely, and because she knew how to influence them. Now she’s launching our migraine group with enormous enthusiasm, because she checked on herself that it works.
Full circle
So, the story comes full circle. It keeps reminding me that a company develops in a spiral: at some point you run into things you already dealt with in the past, except this time you have to solve them at a new technological, scientific or business level, or at a new scale. Right now we’re aiming at a far more specific audience, and far more is thought through in our setup, both in terms of how we do it and in terms of what it could become as a scalable way of looking after yourself.
I can imagine a world where having your own support group around your health, with experts, the latest science and your data, is a standard part of life, available to you every day, 24/7, 365 days a year, for less than the price of a cup of coffee. Instead of just wearing some wearable and glancing at a recovery score now and then with no hope of reducing the number of days you feel like shit.