How Do I Make the Best App Screenshots for the Store

Open the App Store, search for a sleep app, and watch what your own thumb does. You read the name, you glance at the icon, and then you swipe sideways through the pictures. App store screenshots are where the download decision actually gets made, and they get a few seconds of attention at most.

Plenty of teams treat them as a job for the week before launch. Someone pulls five screens out of the build, drops them into a device frame, writes a caption under each one, and uploads the lot. The set describes the app accurately and sells nobody anything.

People decide in the swipe, not the description, and hardly anyone reads that far anyway.

We treat the store page as the first screen of the product rather than the last piece of marketing. It sets an expectation the app then has to meet, and it quietly filters out people who were never going to stay past day one. That second job matters as much as the first.

What follows covers what each store actually displays, the sizes and formats they want, how to choose what to show before anyone opens a design tool, how to write captions people can read at thumbnail size, and how to keep the set working long after launch day.

What the Stores Display and Where Users Actually Look

Apple and Google both pull your screenshots into search results, which means the first image has to do two jobs. It has to win the tap while sitting next to four competitors, and then it has to open the argument once someone lands on the page. Those are different jobs and the same picture has to handle both.

Where App Store Screenshots Appear

On the App Store, a portrait set shows the first three images in search results. A landscape set shows one. Google Play behaves differently again, giving the gallery a strip across the listing and pulling images into search cards depending on the surface. The practical effect is the same in both places. Image one carries more weight than images two through ten combined.

An average visit to a product page lasts no more than ten seconds, and roughly half of that is spent looking at screenshots, according to SplitMetrics. Five seconds is the realistic budget. That sounds brutal until you remember how you behave yourself, scrolling a category at speed, rejecting things on feel.

So the question stops being "what does our app do" and becomes "what can somebody take in during one swipe". Those are very different briefs, and the second one is much harder to write for.

Size, Format and Upload Rules for Apple and Google

The screenshot guidelines on both platforms shift every year or two, so treat the table below as the shape of the thing and check App Store Connect and Play Console before you export anything. Apple in particular has simplified what it asks for, and a lot of advice still floating around tells you to supply device sizes nobody needs any more.

Vertical is the default for phone apps and the thing to deviate from only with a reason, because a landscape set cuts you down to a single image in Apple search results. Games with a landscape play area have that reason. A skincare routine tracker doesn't.

Build the set at the largest size and check it at the smallest. Type that reads beautifully on your 32-inch monitor disappears entirely in a search result on an older phone.

Deciding What to Show Before You Design Anything

A good exercise here is to question every single bit of information you're thinking of putting on screen. Think about the use case. Think about the emotional state someone's in when they arrive at your product page. Does this detail add something for them, or does it purely add complexity to an image they'll look at for a second and a half?

That emotional state is where teams over-promise. A meditation app that leads with a serene stock photo and a promise of calm in sixty seconds is talking to someone who's already relaxed and curious.

Nobody downloads a meditation app because they feel calm. They download it because they don't.

Category choice feeds into this too, because it decides which apps yours gets compared against. Where a product sits plausibly across two categories, we generally advise clients to look at the less competitive one first, build rankings and organic impressions there, then consider moving into the tougher primary category later once there's awareness behind it. Your screenshots sit in a different line-up depending on that call.

Before any design starts, work through this in order.

1. Write down the emotional state someone's in when they find you.
2. Name the one moment in the app that pays that state off.
3. List the three things a sceptic needs answered before installing.
4. Decide which category you're competing in, and look at who you'll sit beside.
5. Cut everything else from the plan.

Building the First Three Screenshots to Carry the Argument

By 2025, around 90% of users don't scroll past the third screenshot, according to ASOMobile. The figure comes with caveats and no stated method, so take the exact number lightly. The direction is right, and it matches how we all use these pages.

So the first three have to work as a complete argument on their own. Everything after image three is for the small group of people who are already interested and want more detail before committing.

1. Image one says what this is and who it's for, in words a stranger understands without context.
2. Image two shows the single moment of value, the thing people actually came for.
3. Image three answers the biggest objection, usually with proof rather than persuasion.

Proof carries more weight than a third feature. A rating, a review line, a recognisable publication, a number of members in a fitness community. Social proof communicates value faster than any description of functionality, because it moves the judgement from "do I believe this company" to "do I believe these people".

If a fourth image is genuinely doing something, keep it. Most sets we see have five, six or seven images that each add a feature and collectively add nothing, because the argument finished at image two and the rest is inventory.

Writing Captions That Sell the Benefit, Not the Feature

Captions get read at the size of a stamp, at speed, often one-handed on a train. Five to seven words is the working limit. Anything longer becomes texture rather than text, and texture doesn't persuade anyone.

The language has a second job, which is managing how someone feels. Are we alleviating anxiety here or adding to it? Are we reducing confidence or enhancing it? "Log your macros, track your deficit, hit your targets" reads like homework to someone who already feels behind. "Three taps and your day's logged" reads like relief. Same feature, different emotional state on arrival.

Export your first screenshot, drop it into a search-result mock-up at actual size, and look at it on a phone held at arm's length. If you can't read the caption, the type's too small and the screen behind it is too busy.

Watch the tone across the whole set as well. Consistent phrasing, consistent typography, consistent spacing from one image to the next, so the set feels like one product rather than six slides made by four people. Inconsistency across a gallery reads as hastily assembled, and people feel that long before they could tell you why.

Avoid the caption that names the feature and stops there. "Smart sync" tells somebody nothing. "Your wardrobe, on every device" tells them what changes for them.

Making Screenshots Match the Brand Promise That Brought People Here

We worked on a mobile app in the health and wellness genetics sector where the brand promise and the product had come apart completely. The packaging and the website were luxurious and aspirational, all about transformation, becoming a better version of yourself, reaching your full potential. The app itself was cold and clinical, purely functional in its copy and its design.

Think about the journey that creates. Somebody buys into the aspirational marketing, orders a kit, does the test, waits, and then opens the app to see their results. The whole thing breaks at the last touchpoint, and the experience turns jarring at exactly the moment it should pay off.

The store page sits in the middle of that sequence, which makes it the place the mismatch either gets exposed or gets fixed. Screenshots that carry the brand's warmth into the product view do real work. Screenshots that show a clinical data table after a campaign about transformation announce the break early.

What Headspace Gets Right

Headspace uses illustration rather than photography, and that's a trust decision most teardowns skip past entirely. A photograph of a serene person implies a standard you're failing to meet. An orange cartoon sun implies nothing about you at all. The artistic style is carrying a large share of the trust, before a single word of copy gets read.

Using Screenshots to Pre-Qualify Users and Cut Early Churn

Getting your store listing right means the people who download already know what the app's about. They've aligned themselves to it as a user. They open it informed, with correct expectations, and that brings the abandonment rate right down.

Which means a screenshot set has permission to put some people off. A fitness app built around heavy barbell work should look like heavy barbell work, so that someone wanting a gentle twenty-minute mobility routine scrolls on. The download you lose there was a one-session user and a one-star review waiting to happen.

A download from the wrong person costs you more than no download at all.

The genetics app we mentioned shows the cost of the opposite approach. Everything upstream promised transformation, so the people arriving at the product were primed for something warm and personal, and what they found was a results table. Expectations set in one tone and met in another produce exactly the quiet early drop-off teams then try to solve with push notifications.

Screenshots fix that cheaply. Show the real interface, in its real tone, doing the thing the app actually does best. The honest version of your product attracts fewer people and keeps far more of them.

Why Referral-Led Apps Still Need a Strong Store Listing

We worked on a gifting and wishlist platform where the client was very hesitant to do anything with App Store Optimisation. The product was social and group-oriented by design, people invite their friends and family into it, so the thinking was that word of mouth would carry the whole thing and store work was money better spent elsewhere.

Our view was that the listing still has to pull its weight, referral or no referral. A friend's recommendation gets somebody to the page. The icon, the copy, the category and the screenshots decide whether they install, and more importantly whether they install knowing what they've got. People really need to understand what the product's about from the very first screen.

Referred users arrive with a vague secondhand description in their head. "It's for organising presents, I think." The store page is where that vagueness either becomes a clear expectation or stays vague and turns into a confused first session.

There's a second effect worth naming. A referred user who looks at a strong page gets a small reassurance that their friend recommended something real. A sparse page with three grey screens undercuts the recommendation that brought them, and the download never happens at all.

Localisation, Dark Mode and Accessibility Considerations

Localising metadata and screenshots can lift downloads by 20 to 40% in global and regional markets, according to The App Launchpad, though the figure's published without a method so treat it as a direction rather than a promise. The underlying point holds up. A German caption machine-translated from English looks exactly as careless as it is.

Localisation goes past language. Currency symbols in the pricing screen, names in the sample profiles, the kind of food in a recipe app's example meal. Those details signal that the product was built for the person looking at it.

Dark mode now sits alongside that. More than half of users run it regularly, which makes a light-only screenshot set look dated to a large share of your audience. If your app supports both, show both somewhere in the gallery.

Check your captions against the image behind them for contrast, and never let colour alone carry meaning. A green tick and a red cross read identically to a lot of people.

Type size is the accessibility issue people miss most on store pages. Caption text sized for a headline on a laptop becomes unreadable in a search result, so set a floor you won't go below and design the layout around it rather than shrinking the words to fit.

Testing Your Screenshots With Product Page Optimisation and Store Listing Experiments

Both platforms now give you proper testing tools. Apple's Product Page Optimisation runs up to three alternative treatments against your current page, splits live traffic between them, and runs for up to 90 days. Google Play's store listing experiments do the same job on the default listing or per language.

The discipline is the same as any experiment. Change one thing, give it enough traffic to mean something, and resist reading a result after four days because you like what it says.

Sensible order to test in, given the first image carries most of the weight.

• The first screenshot, as a whole concept rather than a tweak.
• Caption wording on image one, benefit against benefit.
• The order of images two and three.
• Illustration against interface, where your brand supports both.
• Light against dark, if you support both modes.

Low-traffic apps won't reach significance on any of this, and pretending otherwise wastes months. With a small audience, make the judgement from first principles instead, show the page to ten people who match your audience, and ask them what they think the app does before they tap anything. Their answers tell you more than an underpowered test ever will.

Common Screenshot Mistakes That Cost You Downloads

Most app store screenshot best practices come down to removing things. The same handful of mistakes appear across lifestyle categories from skincare to pet care, and none of them are hard to fix once somebody names them.

• Raw screens in a device frame with no caption, leaving the interface to explain itself at stamp size.
• A first image showing the login screen, which is the one part of the product nobody wants.
• Captions naming features rather than what changes for the person reading.
• Stock photography of delighted strangers instead of the actual product.
• Six images where the argument finished at two.
• A set left untouched after the app was redesigned, so the screenshots show a product that no longer exists.

That last one deserves its own warning. Mocked-up screens showing features still on the roadmap, or an interface tidied up far beyond what ships, create exactly the gap the genetics app had. Somebody arrives expecting what they saw, finds something else, and leaves inside the first minute.

Honest screenshots of a good product beat beautiful screenshots of an imagined one, every time.

Keeping the Screenshot Set Alive After Launch

Store listings rot quietly. The app gets three releases, the brand refreshes its palette, a new flagship feature ships, and the gallery still shows the version from eighteen months ago because nobody owns it.

Screenshot swaps are also one of the cheaper changes you can make. The four-week keyword stabilisation window resets whenever you touch an indexed field like the title, subtitle or keyword field, while screenshot swaps and promotional text edits don't reset it the same way, according to AppTweak. So you can iterate on images without disturbing your ranking position in the way a title change does.

Put a recurring reminder in the calendar to review the full set every quarter, and add "update store screenshots" to the definition of done for any release that changes the interface.

Seasonal updates earn their keep in some categories. A travel app showing winter sun in January, a food and drink app showing the right kind of meal in the right month, a jewellery app leaning into gifting from late October. Small changes, and they make the listing feel tended.

Review quotes shift too. A line pulled from a five-star review eight months ago works harder when it's replaced with a current one, and it tells anyone paying attention that people are still using this thing.

Conclusion

App store screenshots do two things at once. They persuade the right people to install, and they warn off the people who'd have churned by Thursday. Teams tend to design for the first job and forget the second, which is how you end up with strong download numbers and a retention curve falling off a cliff.

The work starts before any design. Decide who you're for, work out the emotional state they're in when they find you, name the one moment of value that pays that state off, and build the first three images around it. Write captions somebody can read in a second and a half. Keep the tone matched to whatever brought them to the page, so the product doesn't contradict the marketing the moment it opens.

Then keep it alive. Review the set quarterly, update after every interface change, and test the first image when you've got enough traffic to learn something real from it.

If you're staring at a store page that describes your app accurately and persuades nobody, that's a solvable problem and usually a quick one. Let's talk about your store listing and what it's doing to the people who find you.

Frequently Asked Questions