RAVER: turning every
small business into a marketing maven.
A redesign of RAVER's AI marketing platform — reframed from a feature-rich beta into a calm, decision-ready product SMB owners actually open before breakfast.
A maven, not a tool.
RAVER V2 helps small businesses craft, schedule, and publish multi-platform campaigns with AI. The beta proved the AI worked — but the experience around it didn't. V2's mandate: tighten the UX, unify scattered prototype features, and ship a product real SMBs would happily pay for.
Powerful AI. Painful product.
Business context
Pre-Series A SaaS targeting SMBs, with Agency and API tiers on the roadmap. Activation and trial-to-paid were both underperforming.
User pain
Signup asked for a credit card before delivering value. Campaign state was ambiguous. Imagery was locked to AI generation only, with no uploads.
Why it mattered
Without a clean activation path, the product's AI advantage was invisible. Fixing the UX was the unlock for growth — not more features.
Listening before redesigning.
We combined 14 SMB interviews, a heuristic audit of the V1 prototype, support-ticket synthesis, and a competitive teardown of Buffer, Hootsuite, Later, and Predis.ai.
Press a button, be done
SMB owners want value before commitment. Every required field before first output erodes trust.
Status anxiety is real
People need to see at a glance: is my campaign DRAFT, ONGOING, or DONE?
Not everyone wants auto-post
Agencies and freelancers often need a downloadable file to schedule elsewhere.
Image control beats image quality
Users want upload, library, and overlays — not only AI generation.
I have 20 minutes between customers. Just give me posts that look good.
I manage 8 clients. I need to export, review, and approve before anything goes live.
I want AI to draft, but I'm the one who hits publish — on my schedule.
Sketch. Test. Sharpen. Repeat.
Ideation.
Two crazy-8s sessions and a 'delete a feature' workshop reduced the scope to what actually moves activation.
Credits, not subscriptions, at the core.
We debated for a week between a flat subscription and a credit system. Credits won — they map naturally to user mental models ("one post = one credit"), let us offer a generous 10 free credits with no card, and gracefully handle plan changes via expiration rules.
One flow. Three states. Zero friction.
A redesigned product built around a single promise: "Tell me what you want to promote, and I'll handle the rest."
Three campaign states
DRAFT, ONGOING, DONE — color-coded pills with strict rules for what each state allows: edit, stop, delete, clone.
Optional everything
Company info, social links, schedule — all optional. Sign up, generate, and download in under 3 minutes.
Imagery your way
Upload, browse the AI library, or add text overlay. Generate 4 more whenever inspiration runs dry.
Schedule or export
Pick a future date, or export the whole campaign as a ZIP with media + CSV for offline scheduling.
Credit transparency
A simple wallet UI shows balance, expiration, and what each action will cost — before the user commits.
Plan changes, no surprises
Upgrade flows preserve unused credits at the old cycle and clearly communicate what's about to change.
From signup to scheduled in 3 minutes.
The numbers loved it too.
What users said
“It finally feels like a product, not a science project.” — Beta tester, week 2 of V2 rollout.
Business outcomes
The credit model raised ARPU 22%, and the download-as-ZIP option unlocked the first agency conversations.
Lessons learned
Defaults are design. Making social linking optional removed our single biggest drop-off — and barely affected post-signup auto-post rates.
What I'd do differently
Run the credit-model concept past 5 users earlier. We spent a week internally debating something users decoded in 30 seconds.