Jeff Halevy weighs in on the biggest challenges facing the fitness industry today, the future of Connected Fitness technologies, and how this will force gyms to up their game.
“I think the fitness industry, as we know it, will go the way of Blockbuster — and that Peloton is not Netflix.”
Imagine the reception this received in 2019 when I rendered my humble opinion to potential investors of my startup, Altis, which sought to create the world’s first AI personal trainer.
At the time, Peloton was still pre-IPO, but it had nearly grown into a billion-dollar company. And 65 million gym memberships represented the lion’s share of a $34 billion U.S. fitness market. I was aware of these facts but had good reasons for my beliefs.
For starters, I had spent over 10 years in the consumer fitness industry at that point — as a gym owner and as a “celebrity trainer” (a term I never liked for a variety of reasons). And nearly everything I did in the industry, I did as an antidote to the pernicious problems that plagued the industry: baseless and faulty, and often cookie-cutter, exercise instruction; and gym memberships driven by use rights instead of value, let alone results of any variety.
As to the former, there are really just two types of consumer exercise instructors; those who instruct a group (i.e. classes), and those who instruct one-on-one (i.e. personal training). What is common to both types of exercise instructor is that neither requires an education in exercise science (it might not be rocket science, but it is a science), nor licensure of any variety.
Just think, if you would simply like to manipulate someone’s musculoskeletal system through massage, a license is required. If you’d like to help someone lose weight with diet advice, a license is required. And if you’d like to enhance someone’s appearance, say by painting their nails, you guessed it, a license is required. But instructing exercise, which not only shares common ground with these licensed professions, but during which an instructor manipulates the body’s musculoskeletal, cardiopulmonary, endocrine and neurological systems, requires no license.
But aren’t these instructors certified? Most often, yes. But before you find comfort in that, just know that almost every certification requires not even a bachelor’s degree in exercise science or any other field, and most necessitate only a few weeks of self-study through an online course.
The fact is most exercise instructors are not even remotely qualified to do their jobs. Have you ever experienced a personal trainer or group exercise instructor, prior to them instructing you to, say, squat, first determine if you even have the requisite mobility in your ankle and hip to do so? If you have, you are in the lucky .01% — cherish that instructor. Perhaps this is why faulty exercise instruction is one of the leading causes for insurance claims in the gym industry.
All that to say that, not only do you often have the wrong people doing the wrong things, but they are doing so en masse. There was a time when at least the “masse” was limited to the 30-some-odd people present in a group exercise class — but now, with the ability to quickly and unforgivingly dose strangers at the other end of a fiber optic connection with digital fitness, the control valve has been blown off the line.
The gym industry is no better. Placing the Average Joe or Jane in even the most well-outfitted gym is as likely to help them produce an effective and sustainable workout regimen as dropping them into Gordon Ramsay’s kitchen is to help them produce a world class Beef Wellington. Likely feeling frustrated, half of all gym members quit their memberships in the first six months.
And the majority of those who maintain a membership visit their gym less than twice per week — less than once per week based on smaller data sets. In fact, that is precisely how the gym industry makes money: by you not showing up. Read that again. Some gyms even outsell capacity by as much as a factor of 20.
Forget the impact of COVID — that nearly 60% of gym members say they will never return to their gyms. What is the long-term viability of an industry that does not want its customers to use its product?
And so, until the early twenty-teens, fitness consumers had roughly three options: go to a place that didn’t want them there to, at best, be instructed by people who were not qualified to do so; receive exercise instruction at home via a DVD; or purchase home exercise equipment and figure it out on their own. For most consumers this is a nearly impossible task even on the uber-simple treadmill, where the only obvious progressive pathway is a linear increase in mileage, time and speed.
None of these was an end-to-end solution in the eyes of a consumer.
Enter “Connected Fitness,” a new category of exercise equipment which, primarily in the interest of delivering streaming exercise instruction, is connected to the internet.
It is not surprising that Peloton emerged as an early winner of this new category. Riding a stationary bike requires practically no skill or instruction (in theory; more on that later). A stationary bike also takes up far less space than a treadmill, elliptical and many other pieces of home exercise equipment. Peloton merged easy-to-use, low-footprint equipment and “instruction,” which is truthfully more like contextual cheerleading (I mean, why else would someone spend more than a few minutes on a stationary bike, let alone make their heart rate uncomfortably high?) — by simply adding a screen and internet connection to the stationary bike. Add in slick marketing, gamification and community, and in hindsight, it’s a damn near perfect recipe for a product that would serve as a category opener, the icebreaker.
And open and break it did. By the end of last year, Peloton had exceeded $1 billion in quarterly revenue. According to the advisory firm Lincoln International, the global digital fitness market ballooned to $27.4 billion in 2020, up 32.6% from $6.7 billion in 2017. Mirror sold for half a billion in cash last year. And this year has brought off-the-chart valuations for competitors, with Tonal and Tempo recently raising over $200 million each, with the former at a whopping $1.6 billion valuation.
However, I am not bullish on any of these options for similar reasons that I was never bullish on the gym industry business model. Peloton, Mirror, Tonal, Tempo and a bevvy of others have virtually the same core formula: a screen connected to the internet that delivers streaming group exercise instruction. Other than Mirror and Echelon, most include and/or are attached to exercise equipment. And then, of course, rather than a membership, a subscription. Semantics.
I am not bullish, but I still bought my mom a Peloton. I figured the contextual cheerleading would help her stick to — and maybe even find some fun in — an exercise habit. I was wrong.
Firstly, getting on and off a Peloton bike is not easy. Pedaling a bike may require no skill, but acclimating to the chore of clicking in and out of the pedals with cycling shoes is awkward, to say the least. My mother found it difficult, but surely once she was all strapped in she would be good to go, right? Wrong. She was bombarded by a screen of video tiles with absolutely no clue where to start. But how could she not know how to get started — couldn’t she just search for a beginner class? Yes; and after searching for a beginner class she couldn’t find a single one that was shorter than thirty minutes, an exercise eternity for someone like my mom. Well, she could always stop the class early after failing to complete it. But at least that failure would come after 10 to 15 minutes, right? Wrong. First, she would need to overcome the nearly immediate sense of failure brought by her inability to maintain the prescribed cadence — during just the warmup.
With my mother feeling defeated, the bike went back to Peloton.
In my humble opinion, this is the disconnect in connected fitness. The product is connected to the internet but is not connected to the user. It doesn’t see the user. It doesn’t understand the user. And it doesn’t personally instruct the user. Connected fitness products are nearly always marketed as “smart” or containing “AI.” Tempo’s device even has a camera in it — but it does not see you — at least not amply to provide form corrections. The oft repeated “knees past your toes” correction that Tempo offers was recently assassinated, and hopefully gone forever, with Knees Over Toes Guy’s 15 minutes.
Intelligence is “the ability to learn or understand or to deal with new or trying situations;” and none of these products, by definition, are currently intelligent. I have nothing but respect for the first wave of Connected Fitness pioneers. But there’s room for so much more. And in a country with nearly 70% of the adult population overweight or obese, and in which entirely preventable lifestyle diseases are the leading causes of death, these imminent improvements couldn’t arrive at a better time.
I believe Connected Fitness 2.0 will address fundamental problems found not only in the first wave of Connected Fitness products — but also the fundamental problems that have plagued the fitness industry for nearly the entirety of its infant life.
So what can we expect?
Connected fitness that is connected to you.
Fitbit, Oura, Whoop and Apple, the dominant players in the health and fitness tracker market, have made it clear that consumers want to know about themselves. They want technology that “sees” them — technology that measures and assesses, that knows how many steps they’ve taken, how many beats per minute their hearts are beating, how long and how well they’ve slept, and even how they are hormonally responding to food.
However, nothing yet quantifies the quality of movement. We have the “how much,” but not the “how well.” And this is arguably the most critical piece of personalized fitness and musculoskeletal health. Whether it’s a cookie-cutter approach in a gym’s group exercise class, or at the hands of an unqualified personal trainer, or at the hands of an internet-connected exercise monolith, a blind eye has been turned toward individual capabilities for too long. Participants are often given the wrong medicine in the wrong dose, a simple formula for orthopedic injury. Can you imagine a doctor practicing group medicine, and prescribing everyone a statin because it’s a “good drug.” Is running a “good” exercise? How about squatting? The real answer is it depends on the individual. For what it’s worth, running, often our first choice when trying to “get in shape,” has an injury rate as high as 90%. Read that again.
This is why I think computer vision is the next fitness tracker. With computer vision we will be able to do for movement hygiene what Oura and Whoop have done for sleep hygiene. We will be able to determine whether an individual’s joints can get into a position where they can absorb and adapt to the stress of a given exercise — the individual’s capacity. We will be able to, further, determine the individual’s competency in movement for each exercise, and we will have the ability to help coach and improve it. Even beyond that, we will be able to determine fatigue levels before and during exercise. Each workout, the right medicine in the right dose.
I could probably even just stop here because I believe, above all else, the future of fitness is in point clouds and pose estimation — GPS for your body.
Connected fitness that connects the dots.
Many health and fitness consumers already feel buried underneath a data dump. We have metrics for steps, sleep, calories, heart rate, HRV and more — but what do we do with this information? Right now, there are few ways to even see all of this data aggregated into a dashboard, let alone an actionable dashboard.
I think the future of fitness sees Oura telling Peloton to take it easy on you today because you barely slept last night and your HRV is down-trending. But APIs are not enough.
Connecting the dots requires intelligence. This type of intelligence is currently deployed by performance directors and strength and conditioning coaches of professional sports teams, and even some MDs and personal trainers. These experts collect and analyze data, from orthopedic screens to movement velocity to heart rate and sleep data, and create robust yet malleable long-term plans that optimize player performance. And while the majority of us will never have millions of dollars on the line based on our on-field performance, wouldn’t it be great to treat ourselves as if we did — as if our body is the single most important asset we have? What a crazy concept.
There is no reason machine learning models (i.e. AI) cannot be trained to think like these experts; there is no reason that the days of unqualified cookie cutter fitness — and the plateaus, injuries and exercise abandonment that accompanies it — shouldn’t be a relic of the past. There is no reason, if we so choose, to not have optimal body function, and the many health and quality of life benefits that accompany it. AI can and should lead the way.
Connected fitness that is connected — well, to everything.
Connecting the dots goes beyond AI and APIs. Connecting the dots also means connecting devices: IoT connectivity. The benefits could range from organic and immediately actionable lifestyle interventions to saving money based on your lifestyle. I wouldn’t mind my car reminding me to park a bit farther from my destination because I didn’t get my 10,000 steps in yesterday. Nor would I mind getting better health, life and disability insurance premiums because my Connected Fitness devices provide direct proof of my preventive health activity — perhaps even one day dynamically mediated by an Ethereum smart contract. And I don’t believe this will create a Big Brother relationship; the data can be aggregated and stripped of any granular details.
To be clear, I don’t think Connected Fitness devices will eliminate exercise instructors or gyms, but I do think it will force both to up their game. And an upgrade in the fitness industry is long overdue.
For me, fitness is not exercise. Fitness, in biological terms, is defined as an organism’s ability to survive and reproduce in a particular environment. With the advent of, first, the Industrial Age and then the Digital Age, our environment has necessitated daily physical activity less and less as each year passes. Aside from not having to move at our desks, with remote work becoming more common and gaining popularity, we don’t even need to go to work. It is now quite possible to live without ever leaving the house, if we are to call that living.
But there are costs to a life of convenience as COVID has recently taught; nearly four of five COVID hospitalizations were among those overweight or obese. The impact of our lifestyles is not always as far off in the future or easily remedied by pharmaceutical interventions as we’d like to believe.
Paradoxically, as our daily movement decreased, our detailed understanding of its necessity drastically increased. Fifty years ago, the concept of running for fitness was new and exercise science as a field of study did not exist. I believe both vectors will continue, so as we become more connected and less mobile, let’s hope that our fitness becomes more connected, too.
