AllDay was Adidas's attempt to do what Nike had done with NTC — build a first-party fitness app that extended the brand into the daily life of the consumer. By the time I joined as Lead PM, we had a functioning app and a growing content library. The challenge was turning casual users into subscribers.
Subscription fitness in 2014-2019 was an emerging category. Peloton was still niche. Mirror didn't exist. The consumer hadn't been trained to pay monthly for digital fitness content the way they had for music or video. We were educating the market at the same time as building the product.
The Hardware-Software-Content Triangle
AllDay was uniquely positioned because Adidas had three assets most fitness apps don't: hardware (miCoach Speed Cell, SmartBall), content (athlete partnerships, training programs), and distribution (global retail, brand trust).
The challenge was connecting those three things into a coherent user experience. Hardware generated performance data. Content used that data to personalize training. The community layer made the experience social.
Getting these three legs of the stool to work together was the hardest product problem I'd faced at that point. Data models had to be flexible enough to ingest different hardware signals. Content had to be tagged in ways that made personalization tractable. Social features had to be opt-in and privacy-respecting for a global user base.
Designing the Subscription Value Loop
The subscription question came down to: what do members get that free users don't, and why is that worth paying for?
We went through several iterations:
-
More content — Didn't work. Users didn't consume all the free content. Offering more of something people weren't using wasn't compelling.
-
Better personalization — Partially worked. Users valued recommendations, but it was hard to make the personalization feel meaningfully different at the paywall.
-
Community + coaching — This worked. The thing users valued most — and that was genuinely hard to offer for free — was access to certified coaches and a community of consistent training partners.
The insight was that people don't pay for features; they pay for outcomes. And the outcome users wanted wasn't "more workouts" — it was "actually getting fit." The coaching and community delivered accountability, which delivered outcomes. That was the subscription.
Localization is More Than Translation
AllDay launched in 14 markets. I was naive about what that meant until I wasn't.
Localization isn't just translating strings. It's adapting the entire product concept for cultural context. In some markets, outdoor training was deeply social and happened in groups — the app needed to facilitate that. In others, solo indoor training was the norm. Some markets had strong martial arts traditions; others were entirely running-focused.
We learned this the hard way. Our initial launch in several Asian markets underperformed because we'd translated the content but hadn't adapted the underlying fitness philosophy. Workouts designed for Western gym-goers didn't land with users whose primary physical activity was badminton or hiking.
We course-corrected by hiring regional content leads who could identify training culture gaps and fill them. It was expensive, but it was the right investment.
What Subscription Taught Me About Retention
Building a subscription product is a masterclass in retention because the economics make it undeniable. Every churned subscriber is a direct revenue hit. There's no hiding behind download metrics.
The single metric I tracked more than any other was day 30 to day 60 retention among new subscribers. If users were still active at day 60, they were highly likely to renew. If we lost them before day 30, they rarely returned.
Everything flowed from that metric. Onboarding, content sequencing, the first coaching touchpoint — all designed to build a habit in the first 30 days. Not an addiction, but a genuine routine that fit into real life.
The biggest lever we found? Getting users to complete their first structured training program, not just their first workout. Program completers retained at 2.3x the rate of one-off workout users. That drove a whole redesign of how we positioned and surfaced programs in the app.