Every study below is real work — synthetic personas walking live product flows, and the outcomes teams shipped on the back of them. Open any one to see the full funnel, per-screen analysis, and the decision it drove.
Diginix, a Dubai AI product studio, was about to hire a $66K PM to hit a four-month delivery. We gave their existing PMs a replica of their target audience instead — they shipped in under three months and saved the hire. Now they're rolling Apriori into three more projects.
We simulated Univest's post-signup activation with 50 personas across four design variants. Users hesitated when the ₹1 activation price stayed hidden behind a “free” framing — surfacing price, proof, and the CTA upfront nearly doubled projected conversion.
100 personas — metro and NRI — onboarded into SuperAstro's ₹1 AI Vedic-astrology chat. The phone-number gate was the single biggest drop-off: users balked at handing over their number before seeing any value.
50 personas through Grip Invest's SEBI-registered corporate-bond flow (up to 12.5% fixed returns). 64% reached invest-ready — the non-linear discovery dashboard held attention, with only a small KYC/PAN dip.
50 personas through Flent's rent-payment onboarding (1% cashback, card float, ₹1.5L landlord insurance). 62% completed — the rental-agreement step was the main friction, costing a 12% drop.
50 personas evaluating Fit Square's pay-per-session gym aggregator. Nearby-gym inventory was the strongest pull; plan selection was where price-sensitive users hesitated, a 10% drop.
50 personas — devout, skeptical, and curious — onboarding into Daily Mantra's Vedic app. 58% reached the paywall; conversion ran on spiritual trust, and the early phone-number ask cost 8%.
Four price-and-framing variants for SaveSage Pro — from a ₹99 lowball to a ₹399 sticker with cashback — tested against credit-card-savvy personas to isolate the tier that maximizes revenue per signup without crushing conversion.
A directional 5-persona read on Loop Health's Healthflex enrollment. 40% dropped at plan selection — the checkout lacked a clear order summary, so users were unsure what they were actually paying for.