what I learned from my failed fitness startup

Author

Collin Stewart
September 27, 2024

Good morning Predictable Revenue community,

Everyone found a different outlet for their time during the pandemic, some learned how to make sourdough, got a dog, or got back into a video game they played in high school (Age of Empires anyone?). I started another startup. My twin boys were 2 ½ and I was still trying to work off my new dad bod so I was focused on something that might help me and others lose weight. We didn’t realize it at the time but even if we were successful, we were never going to accomplish that mission. In this email, I’ll share what I learned about exercise and weight loss along the way. 

TL;DR:

  • Exercise alone does not contribute to significant long term weight loss. You can’t outrun your fork. 

  • The only way to lose weight is to consume fewer calories than you burn. Exercise definitely helps though.

  • The MyFitnessPal (additive model of energy expenditure) might not be correct, if I go on a 1k calorie bike ride it doesn’t mean I get to eat an extra 1k calories and still lose weight.

  • If I want to live long and prosper I need to exercise several hours a week and make sure weightlifting is a part of that routine. 

The goal of Athlon was to help people establish and maintain fitness habits by making it fun. We thought that if we could help people with these two goals, we could help them lose weight. We interviewed 46 people in various walks of life but ended up centering around the persona one of our early participants dubbed “fat dads”. They were dads that used to get regular exercise but fell out of the habit after having kids. Our hypothesis was if we can help people improve their fitness habits we could help them rid themselves of their dad bod. 

Most people we interviewed were tracking their minutes of activity somewhere so we built games that you could “play” by using your minutes exercised as a points system. We had a fantasy football inspired game, a solo adventure where you slowly climbed Everest, and, my personal favourite, one where a group was getting chased by a bear and the person with the fewest minutes each day got eaten. It was really fun to build these silly games and early users seemed to enjoy them but we struggled mightily with retention. 

What I learned

I don’t have a background in exercise science so what I’m sharing is what I learned by reading books, a number of studies, and interviewing professionals. I’ll do my best to deliver a directionally correct version but I know enough about the subject to understand that I have gaps in my knowledge. I’m sharing because I’ve had too many conversations with friends and family about my weight loss journey and most people do not understand the current hypotheses regarding how we burn calories. I wanted to have a post I could point people to if they were interested in learning more and I figured you might be interested as well. 

The problem we ran into was our own ignorance of how calories and exercise are related. We all believed in the MyFitnessPal (MFP) method of losing weight. That is, if my body normally burns 2000 calories and I did 1000 calories of exercise, then I can eat 3000 calories and not gain any weight. I realized our mission might be in jeopardy after reading two books back to back, Exercised by Daniel Lieberman and Burn by Herman Ponzer, both are fantastic reads. 

Burn introduces the Constrained Model of Energy Expenditure, which predicts that energy spent in physical activity will induce compensatory energetic savings elsewhere because our bodies evolved to prevent starvation. See the visual below and note that SleepEE (Sleep Energy Expenditure) is reduced in the constrained model, that’s your body reducing the calories it’s spending on background processes. It's like when your old Windows 95 computer would run out of resources and freeze up, so you'd press CTRL ALT DEL to close all the programs running in the background. When we learned this, we realized our “play games to lose weight” mission was impossible. We tried a few other iterations but slowly lost momentum after this realization.

Exercised was a wonderful read on the subject of exercise, it covered much wider ground and the author has a pleasantly dry sense of humour. I was looking for “the perfect dose” of exercise that we could build our models around and it doesn’t exist. “Given that each of us is an “experiment of one” with different backgrounds, goals, and predictions that change with age, there can be no optimal mixture of exercise type any more than there can be an optimal amount,” Daniel Lieberman, Exercised. So there is no perfect dose that you can apply to everyone but most of us fat dads had similar goals. We exercised so we could keep up with our kids and not die early. These two themes came up in most of our interviews with fat dads. However, he does leave us with some excellent general advice, “Exercise several hours a week, mostly cardio but also some weights, and keep it up as you age” Daniel Lieberman, Exercised.

The interviews always reminded me of this Louis CK quote: 

"When I die, I want to be just skinny enough that people are a little surprised. I don’t want people to go, 'Oh yeah, of course, he died.'"

One guess I’ve made about improving my own weight loss came something I learned in Exercised. Muscle is a metabolically expensive tissue that can account for up to 20% of resting energy expenditure[1]. So if I want to burn more calories while I’m resting then I needed to add more muscle. I can’t say that it did or didn’t work but it at least got me to start lifting weights for the first time in a very long time. Apparently I’m ok working hard just so I can be lazy. And it does appear to be working, at least a little bit.

Back to Athlon

We built an interesting product and had some happy users but we couldn’t figure out retention or a business model that excited us. Any game we built would be easily copyable so there wasn’t much of a moat we could build and none of us had a strong background in B2C or even gaming. We explored different ideas to keep it going but ultimately ran out of steam when covid restrictions eased off and our regular lives picked up again. 

While the startup never really made it off the ground, I did learn quite a bit about exercise and why we lose weight. Here are my biggest health related takeaways from the experience: 

  • Exercise alone does not contribute to significant weight loss. You can’t outrun your fork. 

  • The only way to lose weight is to consume fewer calories than you burn.

  • The MyFitnessPal (additive model of energy expenditure) might not be correct, if I go on a 1k calorie bike ride it doesn’t mean I get to eat an extra 1k calories and still lose weight.

  • If I want to live long and prosper I need to exercise several hours a week and make sure weightlifting is a part of that routine. 

Thanks for reading and let me know if you enjoy "off the beaten path” emails like these. Additionally, if you have any recommendations, books to read, studies to check out, etc… I am still super interested in the subject.

Collin

Sources: 

[1] Energy metabolism: tissue determinants and cellular corollaries. New York: Raven Press; 1992. p. 61–80.

PS - while I was writing this, I emailed Eimear Dolan, a first author on the Energy Constraint and Compensation paper, because I had a few questions. Her replies were very helpful so I’ve added an abbreviated version of our conversation below with their permission.

Collin:
I found your work by following Dr Ponzer's work. I'm a non-competitive endurance athlete and am trying to lose weight by ending each day in a negative calorie balance. However, I frequently go on 2 - 5 hour bike rides where I burn 1500 - 5000 calories.

Do you have any guidance on how I should think about calorie consumption on days where I go on long rides? I believe the CMEE is accurate but I am totally lost when I consider going negative. 

Eimear:
It's a great question. The unfortunate reality is that outside the lab setting (and if I'm honest inside it too!), it is really really difficult to track an individual's energy expenditure and intake, so trying to plan a calorie deficit can be tough. The research shows that in general, energy expenditure tends to be over-estimated and energy intake under-estimated, so often people may think they are in a calorie deficit when they aren't really. This is partly due to difficulty in accurately tracking what we do and eat throughout the day, and partly because the more you train and burn exercise-associated calories, the more compensation may kick in to induce energy savings elsewhere. This can range from simply being a bit less active throughout the day outside of your rides, or in situations of more severe energy restriction, to potentially inducing metabolic suppression. I've attached another paper that our group wrote on the theme of low energy availability in athletes. This one also takes an evolutionary perspective, and the idea for the energy constraint paper that we wrote with Dr. Pontzer initially came while working on this one.

Tracking your calories can definitely be useful, particularly to find where your weak spots are, but I personally am hesitant to trust the calorie data too much, because as I said above, it is notoriously difficult to accurately estimate energy in/energy out. Obviously I don't know anything about your lifestyle, body comp and overall diet, but if you are riding a lot and still struggling with weight then it's probably worth taking a closer look at your diet, both in terms of quantity and quality. Quantity of calories is definitely important, but perhaps the best way to get to an appropriate quantity of calories is by focusing on the quality of what you're eating, particularly on days you do long rides. A diet based on plenty of minimally processed high fibre containing carbohydrates (plants!) + lean protein and minimal ultra-processed foods tends to be satisfying and nutritious, but is also not too energy dense so is less easy to over-eat on (caveat that I don't know anything about your current diet!). I'm currently reading Prof. David Raubenheimer's book "Eat like the Animals" and am finding it really interesting. If you are interested in how evolutionary perspectives influence our eating and activity choices today you may enjoy this book too.

Another thing to keep in mind is that most of the benefits of exercise are independent of its influence on body mass - healthier hearts, better immune function, vitality, improved mental health etc. So if you're riding on a regular basis you're probably in pretty good shape, even if you'd still like to lose a bit more weight.

Collin
I hear you about reliably tracking calorie intake. How reliable do you consider watches with heart rate sensors for estimating calorie expenditure?

Eimear
The data on the accuracy of smart watches isn't great to be honest. They are decent at monitoring things like steps, heart-rate etc, but for total energy expenditure, the estimates really aren't great.

In terms of balancing calories, particularly on days you do long rides, I actually do think that estimating your requirements using predictive equations is useful as it can help to plan your intake. For me though, the trick is to treat these estimates as a rough guideline rather than as a rigid rule, and to have some other indicator that you monitor to let you know if you need to dial it up or down. Body weight is one of these indicators, but my recommendation would be to focus more on other things, like how you feel on the bike, performance stats, how quickly you recover and overall energy and vitality levels throughout the day. These things are harder to define, but ultimately more valuable.