Web design
Morfed — Early-Access Page
Mobile-App Sign-up Site (One-Pager)
Web design
Morfed — Early-Access Page
Mobile-App Sign-up Site (One-Pager)
Challenges
Before launching Morfed, our AI photo-and-video generator, we needed a single-purpose landing page that would:
Capture early-adopter emails without information overload.
Show, don’t tell — a quick demo of the app in action.
Feed real-time analytics into our product-market-fit loop.
That meant embedding a lightweight product demo video inside a phone mock-up, automating the email funnel with Kit (-formerly ConvertKit), and keeping total page weight tiny so Instagram/TikTok visitors wouldn’t bounce on mobile data.
The process
Phase | Key Actions |
|---|---|
Research & Positioning | • Audited top “wait-list” pages (Runway, Lensa, Superhuman).• Ran 8 speed-interviews with creators: “Show me in 10 seconds or I’m gone.” |
Wireframe & Copy Sprint | • Hero: seven-word headline, ten-second MP4 demo auto-loops inside an iPhone frame, single CTA button.• Copy cap: < 140 words total. |
Prototype & Integration | • Built one-pager in Framer.• Embedded web-optimised MP4 (560 KB, 720 p, muted & autoplay).• Kit form drops contacts into “BetaWaitlist” tag; Plausible tracks scroll & click. |
Iterate (A/B Copy Tests) | • V-A hero copy: “Turn Selfies into Cinematic LoRA Art” → 59 % bounce.• V-B: “AI Photos, You-Style. Join the Beta.” → bounce ↓ to 41 %.• Learned Gen-Z creators react better to “You-Style” phrasing. |
The outcome
Metric | Before copy-change | After copy-change | Insight |
|---|---|---|---|
Email sign-ups | 0 (after ~600 unique visits) | 0 (after ~400 unique visits) | Our offer / value-prop still isn’t resonating. |
Avg. time-on-page | 3–4 min (long curiosity scroll) | ≈ 8 sec (quick bounce) | Shorter copy made the “aha” faster—but didn’t trigger action. |
Scroll-depth past hero | 72 % | 28 % | Users now decide in the hero; no second chance to convince them. |
Segment self-ID form | 41 % creators (but tiny N) | – | Sample too small to trust. |
Performance | 930 ms load on 4 G | 680 ms (after compressing MP4 demo) | Speed is solid; problem is message not tech. |
Take-away: Even a “failed” landing page is valuable—because it quantifies disinterest.
We learned that a slick demo + minimal copy isn’t enough; visitors still don’t see why Morfed matters to them.
Conclusion
Data over hunches — Now we know the hero copy needs a stronger benefit (e.g., “Generate branded AI photos in 60 sec—no prompt writing”).
Early signal, cheap fix — Tweak a heading, not a whole app.
Iteration engine — Landing page + analytics lets us A/B a new hook every afternoon until sign-ups appear.
Validates next step — Run creator-focused ads to this page; if bounce persists, pivot wording or audience before spending on development.
The page’s job is not to make us feel good; it’s to show real interest—or the lack of it—fast and cheaply. That’s exactly what it delivered.
Challenges
Before launching Morfed, our AI photo-and-video generator, we needed a single-purpose landing page that would:
Capture early-adopter emails without information overload.
Show, don’t tell — a quick demo of the app in action.
Feed real-time analytics into our product-market-fit loop.
That meant embedding a lightweight product demo video inside a phone mock-up, automating the email funnel with Kit (-formerly ConvertKit), and keeping total page weight tiny so Instagram/TikTok visitors wouldn’t bounce on mobile data.
The process
Phase | Key Actions |
|---|---|
Research & Positioning | • Audited top “wait-list” pages (Runway, Lensa, Superhuman).• Ran 8 speed-interviews with creators: “Show me in 10 seconds or I’m gone.” |
Wireframe & Copy Sprint | • Hero: seven-word headline, ten-second MP4 demo auto-loops inside an iPhone frame, single CTA button.• Copy cap: < 140 words total. |
Prototype & Integration | • Built one-pager in Framer.• Embedded web-optimised MP4 (560 KB, 720 p, muted & autoplay).• Kit form drops contacts into “BetaWaitlist” tag; Plausible tracks scroll & click. |
Iterate (A/B Copy Tests) | • V-A hero copy: “Turn Selfies into Cinematic LoRA Art” → 59 % bounce.• V-B: “AI Photos, You-Style. Join the Beta.” → bounce ↓ to 41 %.• Learned Gen-Z creators react better to “You-Style” phrasing. |
The outcome
Metric | Before copy-change | After copy-change | Insight |
|---|---|---|---|
Email sign-ups | 0 (after ~600 unique visits) | 0 (after ~400 unique visits) | Our offer / value-prop still isn’t resonating. |
Avg. time-on-page | 3–4 min (long curiosity scroll) | ≈ 8 sec (quick bounce) | Shorter copy made the “aha” faster—but didn’t trigger action. |
Scroll-depth past hero | 72 % | 28 % | Users now decide in the hero; no second chance to convince them. |
Segment self-ID form | 41 % creators (but tiny N) | – | Sample too small to trust. |
Performance | 930 ms load on 4 G | 680 ms (after compressing MP4 demo) | Speed is solid; problem is message not tech. |
Take-away: Even a “failed” landing page is valuable—because it quantifies disinterest.
We learned that a slick demo + minimal copy isn’t enough; visitors still don’t see why Morfed matters to them.
Conclusion
Data over hunches — Now we know the hero copy needs a stronger benefit (e.g., “Generate branded AI photos in 60 sec—no prompt writing”).
Early signal, cheap fix — Tweak a heading, not a whole app.
Iteration engine — Landing page + analytics lets us A/B a new hook every afternoon until sign-ups appear.
Validates next step — Run creator-focused ads to this page; if bounce persists, pivot wording or audience before spending on development.
The page’s job is not to make us feel good; it’s to show real interest—or the lack of it—fast and cheaply. That’s exactly what it delivered.
