The paradox of design improvement
Super Unlimited VPN — the #1 VPN app globally by downloads, with over a billion installs and more than a million new daily downloads, almost entirely organic — has spent years trying to modernize its App Store wiki:screenshot creative. The team has tested contemporary layouts, updated color palettes, new content hierarchies, and on-trend design language. The results have been consistent: 80% of the time, the new version loses.
"People just like to see what they were used to seeing," says Tanuj Chatterjee, the app's CEO. "In many cases, we have gone back to the original ones that we had."
The finding cuts against one of the most common instincts in mobile growth: that better design means better conversion. For an app sitting at the top of App Store search rankings, the risk of a major visual overhaul is asymmetric. Disrupting a proven asset is more dangerous than the marginal upside of a slight lift. The team still tests — methodically, one variable at a time — but they've learned to treat the data as the authority, not their aesthetic judgment.
What works at scale doesn't work in a vacuum
The Super Unlimited case study exposes a critical gap between what looks good in isolation and what converts in context. Screenshots that feel dated to a designer or growth lead may carry associations, visual cues, or familiarity signals that new users rely on to make fast trust decisions. At Super Unlimited's scale — where top-of-funnel volume is the primary growth driver — even a 2-3% drop in wiki:conversion-rate from a "better" design can cost millions of installs.
This is not an argument against testing or iteration. It's an argument for humility. The team continues to run experiments, but they've abandoned the assumption that newer equals better. When a test loses, they revert. When a test wins, they roll it out. The process is driven by evidence, not intuition.
The risk profile changes with rank position
For apps lower in the funnel — those acquiring traffic through paid channels, custom product pages, or browse surfaces — creative refreshes often do improve performance. Novelty can break through ad fatigue. Visual clarity can reduce friction. Strong differentiation can capture attention in a crowded category.
But for an app already ranking #1 organically in a high-volume category, the calculus is different. The baseline performance is strong. The audience is familiar with the brand. The creative has been seen millions of times. In that context, a redesign doesn't just need to be "better" — it needs to be materially better, enough to offset the risk of disrupting learned user behavior and visual recognition.
Super Unlimited's experience suggests that at the top of the rankings, the burden of proof shifts. The existing creative has earned its position through data, not opinion. Replacing it requires evidence of improvement, not just aesthetic preference.
What five years of A/B testing teaches
The team's testing discipline offers several takeaways for practitioners:
- Test one variable at a time. Wholesale redesigns make it impossible to isolate what changed the outcome. Iterative testing — changing layout, then color, then messaging — builds knowledge over time.
- Let the data override intuition. If a test loses, revert. If it wins, roll out. Resist the temptation to rationalize poor results or cherry-pick segments.
- Recognize the cost of disruption. For high-traffic apps, even small conversion drops compound quickly. A 2% drop at a million impressions per day is 20,000 lost installs — every day.
- Familiarity is a feature, not a bug. Users process visual information quickly. A layout they've seen before reduces cognitive load. A redesign might be objectively clearer but subjectively unfamiliar — and the latter can dominate the decision.
The broader philosophy: product over presentation
The screenshot story is a microcosm of Super Unlimited's broader growth philosophy. The app's wiki:conversion-rate-optimization-cro strategy centers on product quality and organic growth loops, not aggressive monetization or polished marketing surfaces. The free version is genuinely useful — users can access dozens of server locations, watch a few ads, and use the VPN without restrictions. The paywall exists, but it's easy to dismiss. No dark patterns.
This approach produces a low free-to-paid conversion rate by subscription app standards. The team is fine with it. The free experience generates the ratings volume and return-visit signals that feed the App Store algorithm, which sustains the organic download velocity. "Our top of the funnel is our superpower," Chatterjee says. "Don't mess with it."
The same logic applies to screenshots. A "better" design that disrupts the top of the funnel — even slightly — is a net negative. The optimization target isn't the creative itself. It's the ecosystem that the creative supports.
When to redesign, when to hold steady
The Super Unlimited case doesn't imply that all apps should freeze their screenshots indefinitely. It implies that the decision to redesign should be context-dependent:
- Apps with low organic visibility or low product page traffic have less to lose and more to gain from bold creative tests.
- Apps in highly competitive categories may need frequent refreshes to maintain differentiation and signal relevance.
- Apps with strong paid acquisition funnels can isolate creative performance in custom product pages and test aggressively without risking the baseline organic funnel.
The hardest discipline in app store optimization aso is often doing nothing when your instinct is to act. Super Unlimited's screenshots feel stale to the team. The data says leave them alone. Five years in, the data is winning.