Garmin CIRQA is the headline. But the real story might be elsewhere

SGIE’s read on Garmin’s CIRQA launch frames it correctly as a hardware jab at Whoop: a $199.99, screen-free band with no subscription, landing right as investor media keeps floating a Whoop public listing. But the tension underneath is bigger than one product. Whoop is a subscription business, full stop – there’s no way to buy the hardware without the membership. Plans run from roughly $199 a year up to $30 a month, and the company’s entire roughly $10 billion valuation rests on that recurring fee, not on device margin.

Public markets reward that kind of predictable, recurring revenue with a premium multiple precisely because it’s so easy to model. Garmin, and now Fitbit and Amazfit, are pursuing the opposite strategy: sell the hardware once, no strings attached. Is Garmin really asking wellness buyers to choose between two business models – pay once or pay forever?

That’s the question CIRQA raises, and it gets sharper given Whoop is reportedly preparing IPO paperwork just as two of the category’s biggest names walk away from its monetization logic – timing that could hardly be worse for a story built on subscription durability. It’s also a question SGIE’s piece of news doesn’t chase any further, because it treats CIRQA as a standalone hardware launch rather than one piece of a bigger move Garmin made the very same week.

In the meantime, Garmin closed a second, quieter deal the same week: the acquisition of TrainingPeaks and TrainHeroic, the coaching and training-plan platforms used by endurance and strength athletes worldwide. Read CIRQA on its own and it looks like a hardware jab at Whoop.

Read it alongside TrainingPeaks and it looks like something else: Garmin buying its way into the very subscription layer it just told the wellness market it didn’t need.

Two deals, one week, one strategy

Garmin didn’t disclose terms for the Peaksware Holdings acquisition, but the shape of the deal is clear enough. TrainingPeaks and TrainHeroic bring roughly 120 employees, a large base of coaches and athletes already paying for structured training plans, and a decade of exactly the kind of performance-analytics relationship Fitt Insider had in mind when it described Garmin’s direction as building a “personal health OS.” Layered onto a quarter where Garmin’s fitness segment revenue grew 42 percent year-over-year to $1.75 billion, the timing reads as deliberate rather than coincidental.

That’s the piece missing from a CIRQA-only news. Garmin isn’t abandoning subscription revenue by launching a subscription-free band: it’s relocating it. CIRQA, at $199.99 with no recurring fee, is the low-friction entry point for buyers who’d never sign up for a $30-a-month membership.

TrainingPeaks and TrainHeroic are where the recurring revenue actually lives, one level up the stack, sold to the athletes and coaches who graduate from casual tracking into structured training.

Hardware becomes the gateway; software and coaching become the toll.

Enter Strava: will AI make the toll booth obsolete?

The complication is that the coaching layer Garmin just paid for is exactly the layer AI is starting to eat into for free – and Strava, of all companies, is the one demonstrating it. Strava’s new Model Context Protocol (MCP) connector lets subscribers plug their training history directly into Claude and ask plain-language questions about fitness trends, recovery and pacing – the same kind of insight TrainingPeaks coaches have historically charged for. Other apps have moved quickly to copy the integration.

The irony is that Strava is doing this while confidentially filing its own S-1 with the US Securities and Exchange Commission (SEC), a process it announced in February 2026 with Goldman Sachs reportedly lined up as lead underwriter and a listing targeted as early as this spring. The prospectus, once public, will tell much the same subscription story Whoop wants to tell:

• Premium memberships: $11.99 a month or $79.99 a year (about 90 percent of Strava’s revenue)
• Annual recurring revenue: more than $500 million, reportedly growing around 50 percent
• Profitable since 2020
• Last private valuation: $2.2 billion, with IPO chatter reportedly aiming closer to $3 billion
• Scale: more than 180 million athletes across over 185 countries

That’s the scale being underwritten by a subscription business, while Strava simultaneously hands those same subscribers a free AI coach that competes with the paid analysis coaching platforms exist to sell. If a general-purpose AI model can already tell an athlete whether their easy days are actually easy, what’s left for a paid coaching layer – Strava’s own premium tier included – to charge for?

Some things, it turns out, an AI model still can’t do. Claude can read the data, but it can’t hand out a King of the Mountain (KOM), a Queen of the Mountain (QOM) or a fistful of kudos – those belong to Strava and the community built around chasing them, not to any AI model reading the same numbers.

That community runs at a scale worth sitting with: Strava’s own 2025 Year in Sport report puts kudos given that year at 14 billion, with the number of clubs on the platform nearly quadrupling to reach 1 million total – a volume of social reinforcement no chatbot conversation replicates, no matter how good the training insight is.

The commoditization risk is real for pure data-interpretation coaching – the “tell me what my numbers mean” layer is genuinely at risk of becoming a free feature everywhere – but insight alone was never what kept people opening these apps every day. Engagement is sustained by the loop of posting an effort, watching kudos and comments roll in, and chasing a segment held by someone in your own club: a habit built on social reinforcement, not on sharper analysis.

AI can commoditize the analysis; it can’t commoditize the reason people come back.

That’s the actual fault line – AI compresses the value of analysis-as-a-service while leaving community and gamification intact as the thing actually driving retention, which is exactly why AI on its own isn’t a threat that ends this business model.

Whether Garmin’s newly acquired platforms have enough of that community moat, or are mostly analysis, is the open question.

The cautionary tale: Peloton

Garmin doesn’t have to look far for a preview of what happens when hardware-plus-subscription hits its limits. Peloton built the model years before Whoop scaled it: sell the bike, lock in the membership. It worked spectacularly until it didn’t. Paid Connected Fitness subscriptions in the quarter reported this February:

• Subscriptions: 2.661 million (down 7 percent year-over-year)
• Revenue: down 3 percent
• Net loss: $39 million

Engagement, not just sign-ups, has been the harder problem – the whole thesis of subscription fitness hardware depends on people still opening the app in year three, and Peloton’s numbers say a lot of them stop. The company’s response is instructive. It raised membership prices in October:

• All-Access: from $44 to $49.99
• App+: from $24 to $28.99

It raised prices while simultaneously cutting costs and revamping hardware with AI coaching (Peloton IQ) to justify the increase. And in March it pivoted hard toward a channel it once treated as secondary: commercial placements in gyms, hotels and apartment buildings, built on the Precor acquisition, because the direct-to-consumer hardware-and-subscription loop alone couldn’t carry growth anymore.

Peloton is proof that subscription revenue is not self-sustaining just because the hardware sold well; it has to be re-earned every month, and when engagement plateaus, the business has to go find new surfaces to sell into.

So what’s the actual lever?

Strip away the branding and four possible monetization levers are on the table: hardware margin, software subscription, community and gamification, and platform breadth. Whoop has staked almost everything on the second lever, which is precisely why Garmin, Fitbit and Amazfit are now offering the first one for free – a direct attack on a single-point-of-failure business model.

Peloton shows that leaning on hardware-plus-software without a strong enough community/gamification layer leaves a company exposed the moment engagement softens. And AI threatens to flatten the software-analysis lever industry-wide, which is exactly why the community layer – KOMs, QOMs, kudos, coach relationships, club leaderboards – is becoming the more defensible asset, not a nice-to-have.

That community layer, though, will always belong to Strava, not Garmin. KOMs and QOMs are Strava’s segments and Strava’s community to own – that particular moat isn’t Garmin’s to take, and it happens to be the exact asset Strava is banking its own IPO on.

Garmin’s differentiation has to come from somewhere else: breadth. The TrainingPeaks and TrainHeroic deals give Garmin hardware, software and now genuine coach-and-athlete relationships in one stack, with data flowing across all three – a stack play rather than dependence on a single lever.

Whoop is heading into public markets as a pure subscription business exactly when free hardware and free AI are attacking that lever from both sides. Strava is doing something similar, complicated further by the fact that its own product now gives away, inside Claude, some of what it charges a subscription for. Garmin doesn’t need any single lever to carry its whole valuation – and that, more than CIRQA’s price tag, might be the real story here.

The Playbook with Sebastien Willefert

Strategic thinking for the sporting goods industry

An operator’s perspective on the industry’s most pressing strategic questions. Sebastien Willefert distills two decades of brand, commercial and marketing leadership into digestible, actionable insights. From growth strategy to community leverage, The Playbook translates experience into answers.

Read The Playbook →