Sona

Most of my work is commercial and tightly constrained — products built to ship and sell against a cost and a calendar. I took on this speculative brief deliberately, to work the muscle that day-to-day product work rarely asks for: starting from a behaviour rather than a brief, and designing a whole ecosystem rather than a single object.

The brief set the territory — a sleep and mindfulness ritual for urban professionals who'd tried everything and distrusted brands that over-promise — and asked for a product, an app, and a campaign. I used it to answer a larger question: if I could define a category from scratch, what would I choose to build, and what would I refuse to?

Before any form, I made three bets about how people actually relate to the end of the day — and pressure-tested each one: against my own reasoning, and in unstructured conversations with people I trusted to push back.

The first idea was a quiet record of restful nights. I dropped it before building it. The moment Sona scored your sleep, it would become the thing it was meant to replace — another metric to optimise, another reason to reach for the phone, another way to fail at rest.

So the founding decision was a refusal. In a category whose entire logic is measurement, choosing not to measure is the strategic bet — it's what makes Sona a different proposition rather than a softer-looking tracker. Everything else follows from that one line: the dock has no screen, the app has one job a night and then disappears, and the only thing Sona keeps is a plain record it draws no conclusions from.

Sona isn't a single product, and it isn't a system in the operational sense — it's an ecosystem of four parts — dock, compositions, app, and brand — each owning a moment and passing to the next. I mapped the transition into five stages and gave each part a stage to hold, then hand on.

I started with fragrance capsules and oils, and rejected both — each instantly recast Sona as a supplement or an aromatherapy product. The answer was incense papers in a linen-wrapped box, indexed like a small publication: no oils, no liquids, just paper, scent, and time.

Beyond static screens, I built Sona's nightly flow as an interactive prototype — onboarding, set-up, and the ritual itself, click by click.

Sona is a concept, not a validated product, and I'd treat it that way: before building anything, I'd test the core bet — does designing the transition actually beat optimising the sleep? — small and qualitative, in real bedrooms. The commercial questions, refill economics and retention, are what I'd model before it became a product, not after.

But the part that travels isn't the dock or the scent. It's the method: start from a behaviour the market has misread, identify the moment everyone else is designing around, and decide what to refuse before deciding what to build. That approach isn't specific to sleep. It's how I'd come at any category where the obvious products are all solving the wrong half of the problem — which is most of the categories worth entering.