What five years of A/B testing taught a billion-install app about screenshot design
Super Unlimited VPN sits at the top of global app store search rankings with over a billion installs and more than a million daily organic downloads. The team has spent years trying to modernize their App Store screenshots. The results have been humbling: 80% of redesign tests lose.
"People just like to see what they were used to seeing," says Tanuj Chatterjee, CEO of Super Unlimited. "In many cases, we have gone back to the original ones that we had."
The team tested contemporary layouts, updated color palettes, new content arrangements, and trending design patterns. The data consistently pointed the same direction. The original wiki:screenshot assets — admittedly stale by modern standards — outperformed newer, cleaner alternatives in wiki:conversion-rate metrics.
The risk of disrupting a proven asset
For an app already sitting at the top of search rankings, the calculus of visual overhaul is asymmetric. The potential upside from a marginal lift in conversion is small. The downside risk of disrupting a proven asset that already drives millions of installs is significant.
This finding cuts against one of the most common instincts in app growth work: that better design automatically means better conversion. In mature, high-traffic products, user recognition can carry more weight than aesthetic improvement. The screenshots users have come to associate with the product become part of the brand anchor.
The team still tests — methodically, one variable at a time — but they've learned to treat the data as the authority, not their design judgment. The discipline matters more at scale. A 2% drop in conversion at a million daily installs is 20,000 fewer users per day.
When to prioritize familiarity over freshness
This does not mean screenshot optimization should freeze. It means the bar for change gets higher as an app's visibility increases. Early-stage apps experimenting with wiki:visual-assets benefit from iteration speed. Established apps with high organic volume benefit from caution.
The pattern suggests a few practical guidelines:
- Test incrementally. Single-variable changes to color, layout, or messaging are safer than full redesigns.
- Watch for context collapse. Screenshots that test well in isolation may underperform when users expect continuity with previous store visits.
- Prioritize clarity over novelty. Visual trends change faster than user habits. A clean, readable layout from three years ago may still convert better than a trendy redesign that obscures the product's function.
- Respect the compounding effect of brand recognition. Users who have seen your screenshots before — even if they didn't install — carry mental associations. Changing those assets resets part of that equity.
The broader lesson: conversion data beats intuition
The screenshot testing pattern is part of a larger theme in Super Unlimited's growth strategy. The company operates with a strict data-over-instinct philosophy. That discipline extends to decisions that feel counterintuitive: maintaining a generous free tier that keeps conversion rates low, serving users in loss-making markets, and refusing to implement dark patterns on paywalls.
The free version of the app is robust enough that most users never convert. The team is fine with that. The free experience generates the ratings volume and return-visit signals that feed the store algorithm, which sustains the organic download engine. Conversion rate optimization is secondary to top-of-funnel velocity.
This is not a universal strategy. Most subscription apps cannot afford to leave that much revenue on the table. But for a product with Super Unlimited's scale and cost structure, the trade-off works. The free tier is the growth loop. The paid tier is the margin.
The same logic applies to screenshots. The instinct to refresh, modernize, and align with current design standards is strong. The data says otherwise. For top-ranked apps, familiarity is a feature, not a bug.
What this means for practitioners
If you are testing visual assets creative on a high-traffic product, treat your existing assets as the control with an established conversion baseline. Do not assume a redesign will lift performance just because it looks better.
If you are early-stage, iterate aggressively. You do not yet have the brand recognition that makes familiarity valuable. Test layout variations, messaging angles, and visual hierarchies until you find a format that converts.
If you are somewhere in between, segment your approach. Test new creative on lower-traffic locales or secondary placements before rolling changes to your primary markets. Use the data to inform the decision, not your design preferences.
The gap between what looks good and what converts is real. The teams that close it are the ones willing to let the data overrule their instincts.