The brand's smart scales and blood-pressure monitors were selling well across 12 markets, and the data footprint grew with every unit shipped: millions of readings a day flowing in from devices, a companion app tracking engagement, and a support operation logging issues in its own system. On paper, that data was the company's next big asset. In practice, nobody owned a plan for it.

AI ambition was everywhere and direction was nowhere. Product teams pitched competing ideas for personalised health insights, churn prediction and firmware diagnostics, each assuming a different platform underneath it. Procurement had three separate ML platform purchases under discussion at once, championed by different teams, with heavily overlapping capabilities.

Health-data regulation made every decision heavier. The 12 markets did not agree on what counted as medical data, where it could be stored or how it could feed a model, and no one had mapped those constraints end to end. Any use case green-lit without that map risked being unlaunchable in the brand's biggest regions.

We ran the engagement as four connected workstreams rather than a slide-deck strategy: map the data, map the rules, score the ideas, and leave behind a plan with owners and dates.

The data-landscape audit traced how readings travel from device firmware through the cloud platform, companion app and support desk, documenting what exists, what condition it is in and who owns it. In parallel, we reviewed health-data regulatory constraints in each of the 12 markets, so every candidate use case carried a per-market permissible / restricted / blocked flag before anyone argued for it.

With that foundation in place, we scored all 26 candidate use cases on value and feasibility in working sessions with product, data and compliance leads, set an explicit build-versus-buy position for each, and shaped the result into a three-horizon delivery plan, starting deliberately small with a single tightly scoped pilot the data could already support.