Which feature to ship — and where to place the paywall — decided with 16 creators
A prosumer AI design tool used synthetic creators to prioritise its roadmap and time its Pro upgrade moment.
Palette's product team had four features competing for one build slot and a paywall that wasn't converting. Rather than ship on instinct, they ran three synthetic studies against 16 creator personas — freelancers, founders, content creators, in-house marketers and agency pros.
16
creator personas
4
features ranked
Day 3–5
activation cliff found
1
paywall re-timed
The challenge
Engineering could build one major feature this quarter, and the team was split between an AI brand kit, a template library, one-click resize and team brand controls. Separately, the Pro paywall was interrupting users mid-task and converting poorly — but nobody knew whether the problem was price, timing or the offer itself.
Our approach
The team defined an audience of 16 creator personas across five segments and four incumbent tools (Canva, Figma, Adobe, PowerPoint). They ran a feature & pricing questionnaire, a first-week activation interview to find where momentum stalls, and a UX-research study watching personas hit the Pro upgrade moment and react in the moment.
The results
Feature demand was segmented, not universal — team brand controls unlock the higher-paying in-house and agency buyers, while creators want one-click resize and the most open creatives want the AI brand kit — so Palette sequenced the roadmap by segment value instead of gut feel. The activation work surfaced a day 3–5 cliff once novelty fades, and the UX study showed the paywall converts only when tied to the exact feature in use, with a per-seat team plan surfaced for team buyers. Palette re-timed the paywall to the moment of need and added a team tier.
Research brief
Research goals
Hypotheses
The audience · 16 personas
33
Avg. age
50% F · 50% M
Gender
14
Countries
Freelancer
Segment
Studies in this project
Open any study to explore the full, live results.