Prime 2.0: Ideating a Tier Program with AI

How I got here

Prime is Porter's tier program for its driver-partners, the mechanism meant to reward the most reliable partners and pull the marketplace toward better behaviour. By 2025 its problem was comprehension more than incentive: partners couldn't reliably tell what the tiers were, what separated them, or what reaching the top tier was actually worth. A program nobody can read can't shape behaviour. It's just a badge.

The reader is unforgiving. Low reading comfort, many languages, moved by concrete things: earnings and status, not paragraphs. So the sharp question wasn't "what should the screens look like." It was: what does an abstract tier journey mean to someone who won't read a paragraph to find out?

That's a metaphor problem before it's a UI problem. And metaphor is exactly the kind of open-ended question where falling for your first idea is the trap.

How I thought about it

I didn't want one narrative I'd already fallen in love with. I wanted a wide spread of framings to choose between, generated fast enough that no single one got precious before it earned it.

That's where AI fit: not to make the design, but to be an ideation medium. A way to explore a dozen different meanings for a tier journey in the time it usually takes to defend one. The operating rule was simple: separate divergence from judgment. Let the model produce breadth; keep the choosing human.

What I actually did

Working from the narrative direction the design lead had framed, I used AI to generate twelve narrative directions for what climbing the tiers could feel like, each a distinct metaphor for the same underlying system, not twelve versions of one idea.

photo drops here

The team narrowed the set across two teams, and it converged on two that I then built out into complete concepts:

• The Gemstone direction ("Porter Ascent"). Tier status as a gem you're forming: progression rendered as a material that visibly improves as you climb, with a full per-tier visual-theming spec.
• The Energy direction ("Porter Momentum"). The alternative framing: spark → current → radiance.

Both finalists came out of the AI-generated set, and both were mine to author. I carried the two forward as the decision put to leadership.

photo drops here

The Gemstone direction was adopted, and it won for the same reason it fit the reader. It turns abstract status into a tangible object you're earning: the visual distance between a rough stone and a finished gem says "you are not yet at the top tier" without a word of copy, which is exactly what a low-reading-comfort, multi-language audience needs.

The meta-point I'd make to anyone hiring: the model produced twelve directions in hours, but the craft was choosing the metaphor that makes an abstract system legible at a glance. AI accelerated the exploring. The judgment stayed human.

photo drops here

Where it landed

The adopted Gemstone direction shipped as Prime 2.0 on 11 September 2025. I'm credited in the program's design & research group.

The honest read, and it's mine: as a deep behaviour-shaping program, it underdelivered. On its awareness and comprehension goal, the job the narrative was actually for, it did well.

The program numbers are team-reported, not independently verified:

• Prime's Gold-criterion awareness rose meaningfully after launch.
• Weekly support tickets on the program fell sharply, PAN-India.
• Most partners went on to explore the Prime page after a homepage visit.

What I carry forward: for an abstract system, the concept is the comprehension. AI let me test a dozen ways of meaning it before committing a single screen, and the direction that shipped was the one that said the most with the fewest words.