What Makes an App Popular
Ask a product team what makes an app popular and the answer usually arrives as a feature list. Push notifications, offline mode, social sharing, an AI assistant, a referral loop. The list grows, the build takes longer, and the thing ships with more in it than the competitor it was measured against. Then the downloads land, the charts look healthy for a fortnight, and almost nobody opens it a third time.
Downloads measure how well you marketed. Retention measures how well you understood.
Popularity is a late signal of something small and early. It comes from what a person feels in the first few seconds, whether the thing they were promised matches the thing they got, and whether using it fits a behaviour they already have.
We've spent years pulling apps apart for clients in wellness, travel, property and lifestyle, and the pattern holds. The products people keep aren't the ones with the longest feature list. They're the ones where the design does the emotional work the brand keeps promising in its copy. What follows is how we think about that, with the projects that taught us it.
Downloads, retention and what 'popular' actually measures
Popular is a single word covering four different things, and teams get into trouble by optimising for the one that's easiest to report. Downloads tell you your listing and your paid spend worked. Daily actives tell you something stuck. Revenue per user tells you the value was clear enough to pay for. Star ratings tell you what your loudest users think, which is rarely what the silent majority who deleted you think.
| Measure | What it tells you | What it hides |
|---|---|---|
| Downloads | The store listing and spend worked | Whether anyone came back |
| Day 1 retention | The first impression held | Whether value repeats |
| Day 30 retention | A habit formed | Why it formed |
| Star rating | What engaged users feel | What leavers felt |
What makes an app popular after day thirty
The drop is steeper than most forecasts allow for. In a secondary analysis of 56 mental health apps, JMIR Human Factors, 2022 reports that around 3% of people who download one are still using it after 30 days. That's a category built on good intentions and high download volume, and it still empties out. So the honest question for any team isn't how to get more installs. It's what happens on day two.
Why feature parity does not win
There's a lot of research on this and it's fairly consistent. Matching a competitor feature for feature doesn't get you adoption, and beating them on features doesn't either. What matters is how people feel while they use the thing, what the specific use case is, and the context they're in when they reach for it.
Simon puts the transplant problem plainly. Taking a feature from a product that does something similar in a different scenario, and dropping it into your product which works in a different scenario, won't necessarily make it work. The feature arrives without the conditions that made it succeed. A one-tap reorder button is lovely in a coffee app, where the decision was made months ago and the user is standing in a queue. The same button in a skincare range, where the user is halfway through comparing two serums, interrupts a decision they're still making.
So the useful audit isn't a feature comparison grid. It's a list of the emotional states your users are in at the moment they open the app, and then a check of whether each screen helps or hinders that state. Most feature backlogs we're handed have never been looked at that way.
UX/UI design built around real psychology
We design app interfaces around how people actually think and behave. User research, psychology-driven UX/UI design and technical specs delivered as one complete package.
The first three seconds and the judgement users have already made
Within the first three seconds of opening an app, a person's brain is making fast judgements based purely on what's in front of them. Does this look professional? Is it cheap or premium? Was it made by people who know what they're doing? They aren't asking those questions consciously, and they're feeling the answers anyway.
People feel the answers to questions they never consciously asked.
We saw this working well in a teardown we ran on a meditation app. The visual design was soft gradients and a slow breathing animation, and the animation was already slowing people down before they'd read a word. The interface was doing the calming, rather than the text describing it. That's the mark of good behavioural design. The product behaves like the feeling it sells.
The cost of getting those seconds wrong is front-loaded. Business of Apps puts it at a quarter of users gone for good by the end of day one. You can't recover them with a clever week-two email, because they've already decided what kind of company you are.
When onboarding contradicts the promise the app made
The same meditation app teardown turned up the error that undid all that careful visual work. The app promised calm in under a minute, then asked the user to pick a goal before they'd done anything at all. Sleep, stress or focus. A perfectly sensible-looking screen.
Nobody opens a meditation app for the first time because they're already calm and curious. They open it because they're anxious and need help right now. Asking someone in that state to choose a goal adds a decision, and a decision is friction, which raises anxiety rather than giving relief. The app had promised one feeling and designed for another.
Write down the emotional state your first-time user is actually in, in plain words, then read your onboarding screens back as if you were that person. Every screen that asks them to think is a screen that works against you.
The fix in that case wasn't complicated. Let the breathing animation run, give them the thing they came for, and ask about goals on the second or third session when they're calm enough to answer well. Personalisation is worth having. It's worth having after you've delivered something, so the question feels earned.
Designing around behaviour people already have
On the concierge app project we ran focus groups, and we deliberately included groups of people living in social housing who weren't the target market at all. What we found is that they were having meaningful conversations with each other. Not superficial ones, genuinely deeper connections, and that depth was what produced a real sense of community in their buildings.
That reframed the whole brief. The client's instinct, like most app briefs, was to pull every interaction into the product. Report a fault in the app. Book the lift in the app. Ask about recycling in the app. Our recommendation went the other way. Let the app take the mundane and transactional work, maintenance requests, fire drill procedures, where the recycling goes, and use the space that frees up to encourage deeper human conversation between residents and with the concierge team.
Popular apps tend to sit alongside a behaviour people already have rather than replacing it. The app earns its place by removing the dull parts of something people already do. That's a far easier sell than asking someone to adopt a new habit because a product exists.
Doing fewer things exceptionally well
Travel products are where we argue this most. Offline mode goes into nearly every travel brief, because a competitor has it and because the scenario sounds obvious. The reality is that building it properly is expensive, and building it badly is worse than leaving it out, because a half-working offline map gets used at exactly the moment a person is lost and stressed.
The numbers behind this are uncomfortable. Toptal reports that 88% of users are less likely to return after a bad experience. One mediocre feature can take the whole product down with it. So our advice is to leave offline support out unless it's genuinely critical to the use case, and spend that budget executing the core journey to a standard people notice.
Before adding a feature, ask what you'd cut to pay for it being excellent. If the answer is nothing, you're funding it out of the quality of everything else.
This is the hardest conversation to have with an investor-facing roadmap, because a short feature list looks like a lack of ambition on a slide. It reads very differently in the App Store reviews six months later.
Community and belonging as retention mechanics
The psychological finding from those concierge focus groups was specific. A sense of community comes from familiarity, and familiarity comes from people accumulating small personal facts about each other. Names, jobs, children, dogs, what someone does at the weekend. The moment you start surfacing those data points between people, belonging happens on its own.
Belonging grows from small shared facts, so design should surface them rather than automate them away.
So we rebuilt the product around that. What began as a building directory with a maintenance logger became something closer to a social network, with a social feed at the centre of the experience. Residents could post, ask a neighbour for help with something, and follow updates from the concierge team and local charities. The building management features stayed, as a second layer underneath rather than the main event.
Retention stopped depending on how often a tap washer needed replacing. It started depending on whether anything interesting had happened in the building that week, which is a far more frequent event. Community is the only retention mechanic that gets stronger as the product ages, because the value is held by the users rather than the feature set.
What the App Store listing has to carry
We built an anonymous messaging app that required registration before anyone could see anything. That decision had a knock-on cost we had to design around. With no way to drip-feed value inside the product, everything we'd normally reveal through use had to be communicated before the download, in the listing, the screenshots and the description.
A listing that describes the product honestly does two jobs. It persuades the right people, and it discourages the wrong ones. The people who download it already know what it's about and have aligned themselves with it as users, so they open it informed, with correct expectations, and the early abandonment rate comes right down.
The space you have is smaller than it looks. ASOMobile, 2025 estimates that around 90% of people don't scroll past the third screenshot, so the first three images decide the install.
Treat your first three screenshots as the real onboarding. If someone who has only seen those three images would be surprised by your first screen, the listing is selling a different product.
Why copying a competitor's 'intuitive' flow backfires
A stakeholder tells you a competitor's checkout feels more intuitive, and the request that follows is to copy the sequence of steps. It almost never works, and the reason is worth understanding.
What they're also describing is that they understand the overall purpose of that product, and that understanding bleeds out into how its individual features feel. The flow feels obvious because it sits inside a much bigger product whose point is already clear to them. Strip that context away and the same interaction design can feel confusing or pointless.
You can't lift the same sequence of steps, drop it into your product and expect the same clarity of purpose, because the context is different, the brand is different and the use case is different. A four-step booking flow that feels calm in a hotel group with twenty years of trust behind it feels like an interrogation in a new wellness studio nobody has heard of.
The useful version of that conversation is to ask what the competitor's users understand before they arrive at that screen. Then work out how your own product builds that same understanding, in your own brand, with your own users. The flow follows from the clarity. Copying the flow doesn't produce the clarity.
Measuring whether any of it is working
Most app analytics setups measure whether people did the thing, and almost none measure how people felt while they did it. Both are knowable. Dwell time on a screen tells you whether someone hesitated. Session time against task completion tells you whether the time was pleasant or confused. Sentiment tracking in reviews and support messages tells you which words people reach for when they describe you.
A practical sequence we'd suggest for any team trying to work out whether their changes are landing.
- Write down the emotional state you expect a user to be in when they open the app.
- Pick the one screen where that state is most at risk.
- Measure behaviour on that screen, so dwell time, drop-off and repeat visits.
- Read the reviews written by people who left, rather than the ones from people who stayed.
- Change one thing, then look again after a full retention cycle rather than a week.
The last point is the one teams skip. Retention changes take a retention cycle to show up, and judging a redesign on seven days of data tells you about novelty rather than habit. Hold your nerve and wait for the thirty-day number.
Conclusion
What makes an app popular is a shorter list than the feature backlog suggests. The first seconds have to look like competence, because people feel that judgement before they've consciously formed it. The promise made in the store listing has to survive contact with the first screen. The product has to sit alongside a behaviour people already have, and it has to do a small number of things to a standard that makes the absence of everything else feel deliberate.
Everything we've learned that's worth repeating came from asking what state a person is in when they open the app. The meditation app got its visual design right and its first question wrong. The concierge product doubled in value when we stopped automating conversations and started encouraging them. Neither of those answers came out of a competitor analysis.
If you're planning a new app, or you've launched one and the second-week numbers aren't what you hoped, the fastest progress usually comes from looking hard at the first thirty seconds rather than adding to the roadmap. We're happy to look at yours and tell you what we see.
Let's talk about what your app does in the first thirty seconds.
Frequently Asked Questions
Downloads only tell you that your store listing and marketing spend did their job. They say nothing about whether anyone came back for a second or third session, which is where real value shows up. Retention is the measure that reveals whether you actually understood the people you built for.
Day 1 retention tells you whether the first impression held, and day 30 retention tells you whether a habit formed. Pair those with revenue per user to check the value was clear enough to pay for, and treat star ratings with caution because they capture engaged users rather than the silent majority who deleted you.
Steeper than most forecasts allow for. A secondary analysis of 56 mental health apps published in JMIR Human Factors in 2022 found that only around 3% of people who downloaded one were still using it after 30 days. That is a category with high intent and high download volume, and it still empties out.
Not automatically. Research consistently shows that matching or beating a competitor on features does not drive adoption on its own. What matters is how people feel while using the product, the specific use case and the context they are in when they reach for it.
It is what happens when you lift a feature from a product that works in a different scenario and drop it into yours. The feature arrives without the conditions that made it succeed in the first place. A one-tap reorder button works in a coffee app where the decision was made months ago, but it interrupts someone halfway through comparing two serums.
Map the emotional states your users are actually in at each point of the journey, rather than listing what competitors offer. That tells you which features will land and which will interrupt. A grid shows you parity, but it cannot tell you whether a feature fits the moment your user is in.
Far earlier than most teams assume. It comes down to what a person feels in the first few seconds, whether the thing they were promised matches the thing they got, and whether using it fits a behaviour they already have. Popularity is just a late signal of something small and early.
Not the longest feature list. Across years of work in wellness, travel, property and lifestyle, the pattern is that the products people stick with are the ones where the design does the emotional work the brand keeps promising in its copy. The experience delivers what the marketing claims, rather than leaving the gap for the user to notice.
Related Articles
Apple App Store vs Google Play Store
The apple store vs play store question usually arrives as a budget question. A founder has money...
How app developers can get their apps into the App Store
Getting an app into the Apple App Store or Google Play is one of those tasks that looks simple from...