Garmin's acquisition of TrainingPeaks caught most of the endurance world by surprise, and there has been plenty of reasoning since about why it happened. After reading other people's takes and sitting with it for a couple of days, I think the move makes a lot of sense. At its core this is a defensive acquisition, an attempt to fend off a set of competitive pressures that are all converging on the same place: Garmin's most profitable customers.
The Deal
On July 22, Garmin announced it had acquired TrainingPeaks and TrainHeroic, the default structured-training platforms for endurance and strength coaches. About 120 people join Garmin, financial terms were not disclosed, and both platforms will supposedly keep operating as device-agnostic ecosystems, syncing with Coros, Suunto, Polar, Wahoo and Apple Watch exactly as before.
Garmin didn't disclose the price, but the5krunner built an estimate from TrainingPeaks' published price list: roughly $36.5 million a year in revenue across TrainingPeaks and TrainHeroic, and a likely price somewhere between $75 and $130 million, with a central estimate around $100 million. For scale, Garmin ended last quarter with about $4.3 billion in cash and marketable securities, and generated $469 million of free cash flow in that quarter alone. So this is not a major acquisition for them financially, less than three weeks of free cash flow. But most acquisitions are really about focus and opportunity cost, not the headline dollar figure. And, ideally, about not pissing off the community in the process. I'm talking to you, Strava, and the FATMAP you shut down.
Garmin's Most Valuable Customer
So why did Garmin want this? Garmin has been a hardware-first company, and its most valuable customer sits at the top of the market: the athlete who pays $900 or more for a multisport watch, upgrades every couple of years, and organizes a whole season around structured training. Garmin's own results point the same way: it credited its record first quarter to strong demand for advanced wearables, the premium end of its range. That athlete is also exactly who TrainingPeaks was built for. Not every high-end Garmin owner follows a structured plan, but the coached athlete TrainingPeaks serves sits squarely inside this group, and the numbers around that crowd are striking: USA Triathlon's 2024 State of the Sport report puts the mean income of its survey respondents at $199,000, with average annual spending of $5,631 on the sport, coaching included. Garmin doesn't break out margins by athlete segment, so this part is my read rather than a reported number, but I'm fairly confident this is the one segment where Garmin still beats Apple and everyone else outright, and where a large share of the margin lives.
This vital segment is also under immense pressure. Coros has been eating into the core: it was reportedly the fastest-growing watch brand on Strava in 2025, it took the majority of wrists at UTMB in 2024, and its $249 Pace 4 was named Wirecutter's best overall running watch this year. Amazfit and Suunto keep pushing up from below on price. And in May, Coros became the first wearable brand to connect athlete data directly to ChatGPT and Claude through its own MCP server, with AI-generated training plans named as the next step. The value brands are no longer competing only on hardware. They are walking straight into coaching, which is exactly the ground this acquisition defends.
Then there is Strava, which confidentially filed for an IPO in January at a reported valuation around $3 billion, and which already owns Runna, one of the strongest app-based coaching products around. A Strava that added TrainingPeaks on top of Runna would have owned both the social layer and the coaching layer of the sport, sitting above every hardware brand at once. We don't know whether there was a bidding war, and I want to be clear this part is speculation, but I would guess Garmin was not the only interested party, because TrainingPeaks makes at least as much strategic sense for Strava as it does for Garmin. But I wouldn't be surprised if part of Garmin's motivation was simply making sure TrainingPeaks didn't fall into the wrong hands.
The Apple Parallel
Blocking Strava explains the timing, though. It doesn't explain what Garmin actually wants to build with this, and that is where the more interesting story is. The comparison I keep coming back to is Apple about ten years ago. In 2016, iPhone revenue declined for the first time in the product's history, and the market read it as the beginning of the end of Apple's growth story. Apple's answer was not a better iPhone. It was to change what it sold: monetize the billion devices already out there with the App Store, iCloud, Music, and later TV+ and Fitness+. By the end of 2018 Apple stopped reporting unit sales altogether and started talking about installed base and services instead. That pivot worked about as well as any in corporate history. Services did roughly $109 billion in fiscal 2025, about 26% of Apple's total revenue, and last quarter set another record at $30 billion, up 14% year over year while hardware stays roughly flat. Services also carry roughly double the gross margin of the hardware. The lesson every mature hardware company took from it: when the device stops growing, the installed base is the business.
Garmin is staring at the same wall Apple saw coming. Watches are now good enough that upgrade cycles stretch out, the low end is being commoditized by Coros and Amazfit, and a hardware company running out of hardware growth needs recurring revenue from the customers it already owns. Garmin clearly knows this, because Connect+, launched at $7 a month, was its attempt to build that services leg in-house, and the reception has been consistently negative: no compelling reason to subscribe, features drifting behind a paywall, a product that read as built for shareholders rather than athletes. But there is one important difference from Apple. Apple's services pivot worked because it owned distribution, a toll booth on a billion devices. Garmin has no App Store to tax. Its pivot has to work by owning a relationship instead, and the strongest relationship in endurance sport is the one between a coach and an athlete. That relationship lives on TrainingPeaks, together with what DC Rainmaker points out is 25 years of data on what coaches prescribed, how athletes responded, and what the outcome was. Data of that kind, at that scale, exists nowhere else.
Garmin will use TrainingPeaks to entrench itself deeper into the high-end coaching ecosystem, and over time it will become what Connect+ was supposed to be, bought rather than built. I don't know exactly what shape that takes yet. Maybe bundling, maybe device features that light up with a TrainingPeaks plan, maybe an AI coach trained on that 25-year dataset. But the direction seems clear to me: the wearables contest is shifting from who builds the best sensor to who owns the athlete's relationship, and Garmin just paid to make sure that at the high end, the answer is still Garmin.