Slipstream
Performance analytics for the 3am ranked grind.
A former Valorant coach built a desktop app that records what actually loses ranked games — reaction time, decision latency, how chaotic your mouse path gets when you panic. He had the app, 600 beta users, and a landing page that explained far too much.
Sell to people who already feel the problem. One dashboard on screen, doing its thing. Three numbers that matter. A price. The site says less than the competition on purpose — a 3am grinder trusts a tool that isn't trying to talk them into it.
the gap before you even move.
move, then realize it was wrong.
how much your aim panics under pressure.
three numbers · no leaderboard · no clan · no noise
The first draft had a features section, a comparison table, and an FAQ. We cut all three. The buyer is allergic to being sold to. We replaced the whole middle of the page with a looping HUD and one line: 'your aim isn't the problem. your first 200ms is.' Trial sign-ups tripled in the test.
Ten directions. One shipped.
Every project starts as ten homepages, fully built. The client picks the one that ships. The other nine aren’t waste — they’re the proof we looked before we leapt.
pickedStructured like a post-game scoreboard: round-by-round table of 'why you lost' vs 'what Slipstream found.' The format targets players already staring at end-game screens.
rejectedFull CSS/SVG HUD dashboard centred on screen — corner brackets, scan-line arcs, tabular neon-green metrics, sparkline strip. No photography. The dashboard IS the product demo.
rejectedOne stat at viewport scale — '183ms' towers over everything. The number IS the argument. Nothing else competes with it. Minimal type-only layout.
rejectedThe product pitched as a simulated shell session: slipstream analyze --player=you. Findings print as terminal output. Speaks to the player who lives in a technical mindset.
rejectedLeft half: player portrait with crosshair overlay. Right half: live data panel with a metrics comparison table. Split composition — 'this is about you and your numbers.'
rejectedGiant SVG precision crosshair fills the viewport. Three metrics orbit it at compass points. The crosshair encodes precision, targeting, and control as one visual.
rejectedSix-card grid of sparkline charts — one per tracked metric with trend lines and before/after values. The product is the data. Gaming setup photo as dim background ambience.
rejectedGiant Orbitron '200' dominates the viewport. The copy is the design — stacked text at massive scale with the 'ms' in neon-green. Editorial confidence: no imagery needed.
rejectedPrice card front and center above the fold. $9/mo, five tracked metrics, one CTA — everything you need to decide without scrolling. Sparse layout is the confidence.
rejectedFull-bleed gaming setup photo, strong dark gradient overlay. The founder's testimonial quote sits as the hero statement over the photo, then three stats and a CTA.
Next.js marketing site. Payload for the changelog he posts himself. Stripe for one plan, monthly or yearly. The HUD is a recorded loop, not live data — it reads as live and shipped in days instead of weeks.
brief to paid sign-ups live
trial-to-paid vs. the old explain-everything page
one plan. no tiers, no 'contact sales'
“I told them to make it smaller. They made it smaller twice. Conversion went up both times.”
