Scope the MVP to two mindsets, not four. Research identified four distinct customer mindsets — confident comparison-shoppers, goal-driven explorers, evidence-seeking sceptics, anxious beginners — each needing a different kind of support. Building for all four at once would have diluted the MVP into a generic compromise. We designed deliberately for the two with the greatest untapped potential — Explorers and Beginners — and left the others for a later phase. What was traded away: near-term reach into two identifiable customer groups. What was gained: a first release with a clear emotional register, instead of a lowest-common-denominator tool trying to reassure and convince at once.

Design for hybrid saving-and-investing, not a pure investment push. The safer, more obvious brief would have been "get more customers investing" — a straightforward conversion funnel. Usability testing said otherwise: forcing a binary choice between saving and investing created exactly the fear the whole project was meant to remove. We designed the choice stage around a blended mix — keep the safety of saving, add a controlled amount of growth — rather than positioning investing as saving's replacement. That meant giving up a cleaner "investing" pitch in exchange for a framing customers could actually accept.

Bet the MVP on the calculator, not a new journey. With six workstreams on the table — including an AI coach — the tempting move was a brand-new flagship journey. The data argued for something less glamorous: the existing calculator was already the estate's most engaging tool, yet only a small fraction of nearly six million yearly visits ever reached it, and it lost nearly half of those at the first screen. Redesigning it became the lead MVP vehicle — trading the novelty of a new experience for day-one reach and a clean baseline to measure behaviour change against.