Radu Popescu
Founder of Gaidd · Growth design for self-serve SaaS
I run a performance-based conversion practice for product-led SaaS. Companies bring me a funnel that isn't converting, and I diagnose where it leaks, design the experiments that fix it, and get paid on the lift that actually holds — not on the hours spent producing it.
The work so far
- 6,204
- Screens audited
- 59
- Split tests run
- 12,042
- Users converted
- 27
- PLG scale-ups served
- 201
- User journeys mapped
How the work runs
Every engagement includes the same six things. There are no tier unlocks and no upsells — the difference between a small client and a large one is the price per point of lift, not what they get.
- 01
Behavioral MAP
Every step of the funnel mapped against session recordings and event data, to surface exactly where users drop and why they drop there.
- 02
Funnel UX Audit
The activation flow reviewed against established conversion principles — friction, confusion, and missing trust signals flagged individually.
- 03
Data-First Check
Each hypothesis validated against your actual usage data before anything gets designed. No gut-feel guessing.
- 04
Sprint Roadmap
Experiments ranked by expected impact, so the team always works the highest-leverage bet first.
- 05
Hi-Fi Design & A/B Setup
Dev-ready Figma variants your engineers can build the same week, with flags, splits and variants configured inside your existing stack.
- 06
Analysis Into the Next Sprint
Every test measured against a real baseline, and the finding fed straight into the following round rather than filed away.
How I think about this work
Paid on outcomes, not hours
The fee is charged per +1 percentage point of absolute free-to-paid lift, and nothing is charged until a validated experiment wins and the lift holds. If the work doesn't move the number, it doesn't get invoiced.
Behavior first, pixels second
Most conversion problems are not visual. They are a user being asked to do too much, too early, before they have seen anything worth the effort. The design follows the behavioral diagnosis.
Turn work down that can't be measured
Below roughly 250 signups a month, an A/B test cannot reach significance in a sensible timeframe — you would be paying to measure noise. Products in that position get told so instead of sold to.
What clients say
“I have 0 doubt that anyone who hires Radu will see a big ROI. If your numbers are not where they should be, Radu is the person to diagnose and improve your UX/UI And PLG Activations.”
Ilija Stojkovski, Chief Revenue Officer at HeyReach“The 'Aha!' moment was happening in the sales call. Radu helped us identify our real Aha moment through data, reverse engineer the activation Journey and help self-serve users find value in the first session”
James Berry, Founder at LLMrefs“The Audit was Insane! We have 2 UI and 3 UX Designers. None mentioned the stuff you said. And they're obvious!”
Immanuel Rajadurai, Founder at ZooPark“You're a hard to find person. I rarely see someone who understands user behavior & psychology as you do. You're a natural.”
Michael Reimer, Fractional Design Lead at SalonScale“Radu did a journey teardown of our onboarding and activation flow. The full UX, identified the 'Aha!' moment, and pointed out friction I'd stopped noticing.”
Martin Markov, Co-Founder & CEO at ClearCueWho this is for
Live SaaS and AI products with a self-serve funnel — free trial or freemium — doing roughly 250 or more new signups a month. Pre-launch products and sales-led motions without a self-serve tier aren't a fit, and I'll say so rather than audit a funnel that can't be measured.
See the full pricing ladder, read the case studies, or book a free growth roadmap.
Start Today and Launch
4 A/B Tests in 14 Days
Deposit €2,000. Launch 4 A/B Tests in 14 days, and get your €2,000 back once they're complete.