About this episode
You do not need an Apple Watch for your wedding day. Put it down. You've made your decision. You do not need to track your heart rate — because you're not running anywhere.
But here's the thing. I think measuring all of this and getting monthly blood tests is cool. It's interesting. But I don't think it works for real people. I think there's a subset of super-responders — people it benefits tremendously. But I don't think it works on real people. People like me. Who are out of shape. Lazy. Unhealthy. And it's not because we don't have the information — it's for a whole host of other reasons that I don't think wearables actually help with.
I argued about this with Maitham Dib, the founder of Vital, a startup that's gone through Y Combinator — basically the Harvard of startups — and has raised $2.9 million from the likes of Point Nine and Harry Stebbings' 20VC fund. His startup is quite cool. He's building an infrastructure layer for all of this wearable and lab data. Loads of other startups and big companies want to integrate wearable and lab data — from the Apple Watch, Whoop, the Oura Ring — but it's really hard and takes a lot of development time. Vital basically makes that possible for them overnight. It's quite similar to Stripe in the payments world.
In this conversation
- A practicing physician opens the episode by calling an entire venture-backed industry bullsh*t — to the face of the founder building it — and the two spend half an hour genuinely trying to win the argument.
- The real crux of the wearables debate: it's not "information is power", it's whether software built on top of the data can actually change human behavior — Maitham cites a metabolic-health customer whose users lost 30% more weight just by connecting a device.
- Why building Vital is brutally hard: standardising HRV across hundreds of devices with missing docs, and integrating with antiquated labs where you have to physically visit and win over the lab director — the "schlep" that becomes the moat.
- A concrete playbook for enterprise sales as a young unknown founder: anchor on your champion, map the stakeholders, and read the inner politics — like the team whose in-house build you're about to replace.
- The 1,000-IQ move behind doing YC as an infrastructure startup: hundreds of other startups in the batch become your first customers — plus why naivety is a genuine founder superpower.
Transcript AI-generated
Let me give you a controversial view related to your industry. This whole thesis around lots of health data being super, super useful — especially in younger, mostly healthy populations, using it as almost preventative or screening —
Or just tracking data for the sake of it —
— and that this will give you some big unlock. I think it's complete BS. I think any conclusion you'd draw from all these wearables — or even a lot of these lab panels — as someone who doesn't have an acute medical need, who's just kind of interested... that ranges from CGM to heart rate variability to VO2 max to even basic blood panels. I just don't think it's that useful. Because if you go down the Bayesian decision tree, the end result is always going to be: don't eat crap, drink less alcohol, don't smoke, exercise more. So I view a lot of the activity in this space as — I don't want to say grifters — but I just don't think it's actually benefiting anyone. I think it's cool, but I don't think it benefits people. What do you think?
I think there's nuance. That's what I'll say. So — what are the three most important lifestyle changes you can make? It's exercise. There's a study released recently showing exercise is the most important thing. And everyone's agreeing — exercise is so important to human health. So: exercise, nutrition, and sleep. They're probably the three biggest, most important factors. Would you agree?
Yeah, I'd completely agree with that.
So the question is: can wearables help us adjust that behaviour? And I think the answer is yes — because I've seen it firsthand.
Can we get people to exercise more using wearables, and by building software that incentivises exercise? I think yes. One of the coolest things about the Apple Watch — it's so basic, but it works — is the ring mechanism. Closing your rings at the end of the day. For me, that incentivises me to go and make sure I get 10,000 steps. I'm often just sitting on the computer all day. But when I get to the end of the day, it reminds me to stand up — it's actually getting me to change my behaviour.
And if you look at CGMs — the whole CGM movement has brought attention to the fact that sugar is so bad for you. It's like you and me — we probably live in a bubble where it's so obvious, we hear about it all the time. But a lot of people just don't realise. It's so easy in the tech bubble to think, this is so obvious, what's special about this? But giving people the ability to see the actual effects on the body when you eat a bowl of white rice or a ton of candy —
— and seeing the effect, and then feeling the effects as well — I think it does cause you to change your behaviour. Versus without that information, people just carry on. Sometimes people just feel fine. I have friends who still drink three bottles of Coke a day, and I'm like, that doesn't seem good for you. And he's like, I feel fine. I'm pretty sure if he put a CGM on himself, he'd see the glucose spike, and he'd start thinking: okay, why am I feeling crappy during the day? Why, when it comes to lunchtime, am I feeling pretty bad? Is it because I'm drinking a bottle of Coke?
I've seen people firsthand change their behaviours by measuring things. And I think that's so important. Same with sleep. I see both sides, but I still sit on the side where measuring these things, and being able to bring them back to a medical professional who's armed with the data and can actually explain to you why this is happening — that's better than the current system. What we have now is people don't know, and sometimes it has really dire consequences that could have been prevented had you measured, seen change over time, and gone and got the opinion of a professional.
So I agree with about 50% of what you said. But I'd actually agree with one of your points — the core thing being behaviour change. And I'd go further. This thesis around information being power and helping people change their behaviours — I can appreciate there's a shorter feedback loop. You can see the consequences of your actions today rather than in 20 years. So I totally get that.
But I had this experience working with vascular surgeons. A lot of their day job is people coming in who are smokers with vascular disease, and ultimately they have to get limbs chopped off. So they spend a lot of their days in the operating theatre chopping people's legs off. It's a really grim thing to happen to someone. And the funniest thing is a lot of them would then go on smoke breaks — which was so funny. These people who are seeing all this information still smoke. And that really convinced me that information, or data, is in some ways almost useless. The most important thing is the identity change, or the inner psychological work, or whatever goes into it. That really started this thesis that more data isn't necessarily useful. Maybe I'm just trying to be contrarian, but — more data isn't necessarily that useful.
“It's not more data. It's data with the right software that incentivises the correct behaviour. Apps built on top of this data can actually change behaviour based off wearables data.”
Maitham
Again — it's not more data. It's data with the right software that incentivises the correct behaviour. Does that actually cause change? Does it change behaviours in humans? I think apps built on top of this data can actually change behaviour based off wearables data.
That's what's key. If we agree with that — what do you think on that point specifically? Do you believe software built off this data has the power to change human behaviour?
Definitely in a subset of the population.
Okay. So the question is — are you saying that subset is the wrong subset to go after?
Oh, that's it. I didn't think of that. That's an interesting point. Maybe the people you need to target wouldn't respond — the people with the biggest need wouldn't respond as much. But I'd predict maybe 20 to 30% of people will respond to this data, this gamification, this shorter feedback loop. That'd be my guess.
What we've seen from customers is that by integrating wearables data — for example, we have a metabolic health company, one of the largest in the App Store, focused on intermittent fasting and losing weight — people who integrated their wearables data lost, on average, 30% more weight. By integrating their wearables data. So for me, that's one nugget that shows there's actual value in integrating data from these devices —
— because it can actually incentivise certain behaviours.
The question is: who's actually using these weight apps? Is it the people in most need of losing weight? That's a different question.
Ultimately, developers need to find a way to build applications that appeal to the right audience. But the fundamental question is: does data help incentivise behaviour change? I think we can both agree it does.
What's the big moonshot opportunity for Vital? Clearly there's something about you being the API for bigger and bigger companies — everyone in the world who needs it. But are there any rogue opportunities, any cool vertical or horizontal, something else I'm not thinking of that you could do?
We've got some interesting stuff coming out — I can't mention it. But in terms of our roadmap, the general aim of Vital is to get this data, in a synthesised fashion, back to doctors. What's really cool about Vital is that, unlike other companies, we're very much focused on healthcare, and on making sure people can consume us almost like a diagnostic-as-a-service — where, if you have any sort of condition, you can use our API to essentially work out whether someone is exhibiting a certain symptom or likely to exhibit a specific disease or problem in the future.
We've got over a million people now physically connected on our platform in terms of devices. We have billions of data points. What's going to be really cool is seeing trends across continuous data and lab data, and being able to do predictive analytics based on it — I think that's going to be a real superpower of ours. And then bringing those synthesised, actionable insights to doctors. That's going to be key for us.
There's a whole movement of lifestyle and functional medicine doctors in the US —
— and I think that's an interesting space for us, because those types of doctors typically look at healthcare holistically, versus a primary care provider in the US. The functional medicine doctor, the lifestyle doctor, will look at things as a whole and try to understand what is wrong with this patient — versus, they're experiencing this acute symptom and we just need to treat them. So from our point of view, that space is really interesting. Over time, we'll be getting that data and the actual insights to the right people — which, in our eyes, is the functional medicine and lifestyle doctors, who'll actually be able to do something interesting with it.
The point around integrating wearable and lab data, and then maybe some subjective experience of how someone's feeling and the clinical exam — that'll be super interesting. Because there's a chicken-and-egg problem right now. Someone comes in with a wearable and says, "My HRV dipped by 5% on this day, and my CRP went from 3 to 4 — what does that mean?" And no one knows, because it's just never been seriously studied. And it's kind of annoying, because people come in with an expectation, you don't know what it means, and you don't have any evidence to give them any good insights.
So I wonder if you'll play a role in that. That would be super interesting.
Yeah — what if you could literally say, this is raised, and this is why, and these are the steps required? It's like a diagnostic-as-a-service, and being able to embed that information to the right people is so important.
That's what's most interesting — our sort of secret sauce is the data across multiple modalities: the continuous data as well as the lab data, which companies often don't have. They segregate the two things. To truly do this, you need both — and then eventually be armed with the medical history of the patient.
Can you talk about naivety as a superpower?
“One of the most powerful things as a startup founder — and Paul Graham writes about this all the time — is being naive. Often you see all the holes in the idea before you've even started, and you talk yourself out of it.”
Maitham
Yes. One of the most powerful things as a startup founder — and Paul Graham writes about this all the time — is being naive. Often when you're overloaded with information and thinking about starting a startup, you see all the holes in the idea before you've even started, and you talk yourself out of actually starting it. Being naive about an industry, but approaching it in an experimental way — treating the startup as a series of experiments where you have a hypothesis and you test it, but test it in real life. I'm going to get my product into the hands of users, see how they react, learn from that, and adjust my initial hypothesis.
Having that naivety — just saying, okay, I'm going to launch it, my idea is probably bad, but who cares, I'm going to try it and learn from it — is a superpower. Versus, if you knew too much about the industry or the specific problem, sometimes it can prevent you from taking action. So being naive is so important as a startup founder — treating startups as experiments, being malleable. Don't be so dogmatic about your initial idea. You need to change along the way as you learn more about the industry.
Some businesses look really simple. Vital is one that looks really hard. I can't imagine how difficult it is to get hundreds of thousands of wearables to operate with Vital, then include hundreds of different labs with their different measurements and standards, and put all that together into a nice API that other companies can access with one click. How do you start something like Vital? What do you do on day one? It seems like such a big beast to tackle — and it's not terribly useful for anyone if you're only incorporating three devices. You need to offer a wide range. So on day one you're probably not that useful, right?
That's true. That's what was hard about the product we were building — there was a time we just had to spend building. And then, as soon as we had about five or so devices, we started approaching companies. The thing that was interesting about our strategy: initially, when we didn't have many devices, we focused on companies that only wanted integrations with certain vendors or certain devices. For example, there's a great cycling app — they were really interested in integrating data from two devices at the time, I think Hammerhead and Zwift. So we said okay, let's build this for this company, and see if they find use in it. And there were other companies focused just on diabetes and metabolic conditions — okay, let's integrate with CGMs and BGMs, and try to sell the API and see what they say. So at the start, Vital looked at certain verticals, broke down each one — what are the most important data sources for that vertical? — and that's where we went with it.
People often don't realise how hard building actually is once you start integrating hundreds and hundreds of devices. Standardising data across them is extremely complex. Very painful. You think documentation is there on how they've calculated a certain HRV — you think there's documentation on which calculation method they used, whether it's RMSSD or SDNN — and it's not. You'll have to figure it out. Or there's missing documentation, or APIs changing all the time. It's really hard.
And then the lab space makes it even harder. As you build more integrations with labs, there's more of a human element. With devices, at the end of the day you're reading data off essentially a small machine, and they have a way to interact with it. With labs, it's a lot more difficult. It's more reliant on personal connections. It's a very antiquated industry, so you still have to go visit these labs, see what they're doing, speak to the lab directors.
It's more of an old-school integration, if I can call it that. And that's what makes it hard. So people don't realise, when they're first starting out, that we've gone ahead and done all that hard work. Paul Graham calls it the schlep. We've done all the schlep of integrating with these different labs and devices — and now people can easily use one API to get that information.
So what are the pros and cons of doing an enabler play, an infrastructure play? You're kind of similar to Stripe in payments — take all the hassle of dealing with payment stuff, here's one line of code you put into your website, and people can have a buy button without the hassle. Similarly, you're not the main thing, not the big hot startup directly doing the care. You're enabling other startups to do all of that.
One of the biggest pros of an infrastructure play is just seeing the number of companies that use your API for different use cases — ones you'd never even have thought of. That's super rewarding. From a business point of view, being infrastructure — it makes it hard —
— you're such a vital part of a company's process. If you're picking a specific data provider, especially a health data provider, you're relying on that company staying around for a while, because it's a big investment — you're basing everything on them. So sales cycles tend to be longer; that's a negative. But once they're actually integrated, you shouldn't really churn that much. You should have negative churn, because you're such a vital part of the infrastructure.
What have you learnt about getting enterprise sales? Because, from a cynical point of view — you rock up, you're a young guy, it's a startup, and if they use you in their infrastructure that's pretty foundational stuff. It's going to be a big problem if you die one day — and statistically, most startups do die. So they're thinking, who's this cowboy? I don't want anything to do with this. And I'm guessing a lot of these decision-makers might be significantly older than you. I can't imagine going into that room and getting a huge, significant contract. What have you picked up about how to gain that trust and get those sales?
“That's the hard bit of enterprise sales — mapping out who the stakeholders are, understanding the inner politics happening within the organisation, and how your solution being brought in influences the company.”
Maitham
You've got to treat enterprise sales the same way you'd treat any business problem — even an engineering problem. You have to work out who the stakeholders are. You look at the first person who did the inbound, who was interested in the company, or whoever you reached out to. You anchor on who your champion is within the process. For example, maybe the CTO reached out — he's interested. He'll be your champion, the person really vouching for your service, because he made the initial reach-out.
When it comes to enterprise, sometimes it's not the CTO. Having a CTO as a champion is great, because he ultimately has the buying power. But sometimes it's a PM, or a developer who's brought you in — and then it's the opposite, you've got to work your way up and understand who the actual decision-maker is. Is it the CTO? The PM? But I'll work with that CTO scenario. Once you've got that champion, what they'll typically do is introduce you to different business units or teams within the company. They're your cheerleader. But at the same time, you need to understand how these business units operate — who do I need to influence in this specific team?
You also have to understand the inner politics. Why have they brought me in? What does this mean to that specific team? For example, if they brought our solution in to replace something they'd built in-house, you have to understand how that person is feeling — their previous work is being wiped away, and now they're just going to use an off-the-shelf solution instead. Which is great for the company, saves cost. But sometimes you've got to deal with that inner politics.
That's the hard bit of enterprise sales — mapping out who the stakeholders are, understanding the inner politics happening within the organisation, and how your solution being brought in influences the company. And then, it's an enterprise deal, so it's about actually meeting these people in person, flying out, seeing them. But the hard bit is really just mapping out who the stakeholders are and understanding the inner politics. What does this mean to the company if they were to use our solution? What other problems can we fix for them? Just being a great listener helps a lot in enterprise sales, but also being proactive about understanding the different dynamics in that organisation.
So you went to the Harvard of startups, Y Combinator. What were the big lessons you picked up there?
Launching quickly. YC hammers it into you — just launch. And that's so important. How many people — and this again comes to, maybe it's your superpower — just launch your first idea. It's never going to be the best idea. It's never going to be the idea that gets you to IPO. It just won't. And it doesn't have to be perfect. It can just be a problem you're fixing for yourself, or something you have strong conviction on. But the most important thing is you get it out there and test. Get it in the hands of real users. Test. Iterate. That's the most important lesson from YC.
Launching quickly, speaking to customers, and focusing on product. Everything else you can kind of ignore until you've got that solid foundation where you're getting customers repeatedly. Just launching quickly, speaking to customers, and making a product that people want. That's their tagline, and it's true.
What about the benefits of the network, just being in that space? There's always this belief around wanting to be the dumbest person in the room, or buying the worst house in the nicest neighbourhood — that always seems to be the hack. And secondly, it seems like a thousand-IQ move from you to do YC — because a startup that provides a service other startups can use, going into an incubator with hundreds of other startups, they can be your first customers, right? So what about the network / being-in-the-right-room aspect, and that thousand-IQ move of getting your first customers through YC?
The network's one of the best things about YC. You have access to a lot of startups, and also companies that are no longer startups — really huge private companies like Stripe. If you wanted to email the CEO and ask them something — obviously not everyone does this — but if you have a problem that's important for that company, you can do that.
And often, having YC behind you, they'll take you seriously — versus, if you were just some random person emailing, you'd often get ignored. And then having access to people one stage ahead of you, where you can ask — okay, maybe we're at seed level, you're at Series A — they'll have gone through the same problems, and having that resource is so powerful.
The second point — yeah, you're right. Because YC has such a wide range of startups, you can sell into them. But the downside: people always over-glamorise selling to startups. Selling to startups is good, and we sell to a lot of them. But they also tend to be some of the most demanding customers, and you can't charge startups a lot of money, because they're growing. You want them to see value in the product, and hopefully they'll pay for it, and they will — but you can't charge them insane amounts. And often you're hoping that startup will eventually succeed and grow to be something pretty big.
Last question. Have there been any habits, or ways of approaching things, that have been helpful for you?
I'm a first-generation immigrant. I always don't like spending a lot of money.
So I artificially set myself constraints — okay, we've raised a bunch of venture money, but at the same time, I don't like to spend a lot of money unless we really need to. And I think keeping the team small until you've truly found product-market fit, and you truly can't handle it — you've truly experienced pain, and the ultimate limiting factor on your growth is people. That attitude of being first-generation, not spending a lot, has helped us. And sometimes it can hinder you, because you're thinking, maybe I don't want to hire more people until we close more deals. So you've got to be careful about managing that balance.
But ultimately — that's the thing that's been ingrained in me as a first-generation immigrant.
Thanks so much for listening. If you've made it this far, you're probably a superfan — or you've got nothing better to do. In either case, we'll probably be friends, so you can email me at hi@musty.io. That's hi at musty, M-U-S-T-Y dot I-O.
I lose money on every podcast I make, and I'm pretty broke, so if you want to help, there are two things you can do. The first: if there's anyone in your life who you think is equally as nerdy as you, please send them an episode — this one, or another you think they might enjoy. Secondly, leave a review on YouTube, on iTunes, on Spotify, wherever you heard this. That helps a lot. Thank you so much, and see you next episode.