What to Put on Your App Store Screenshots (and What to Leave Off)

Nobody reads your App Store screenshots. They glance at the first one, maybe half of the second, and decide. That is the whole interaction, and most listings are built as though the opposite were true.

Which is why the advice that matters is not really about design. It is about what you choose to say in the two seconds you get, and what you are willing to leave out so that the two seconds are not wasted.

The first screenshot is doing most of the work

Apple shows up to three screenshots in search results. Google Play shows two or three depending on the surface. Everything after that is only seen by someone who has already stopped scrolling and started swiping, which means they are already interested and much harder to lose.

So the first image is not an introduction. It is the pitch. Whatever single thing makes someone want your app should be on it, stated plainly, large enough to read on a thumbnail roughly the width of a thumb.

A test that works surprisingly well: shrink your first screenshot until it is about 150 pixels wide and look at it. If you cannot tell what the app does, neither can anyone scrolling past it. Most sets fail this on the first try, usually because the caption is too long and the device frame is too small.

Captions: short, and about the person reading them

Two to six words. Beyond that you are writing for nobody, because the text gets smaller to fit and the smaller it gets the less likely it is to be read at all.

The other half is what the words are about. Feature names describe the app. Benefits describe the reader. The second converts better, and it is not close.

Notice that none of the right-hand column mentions a feature by name. That is deliberate. The feature is visible in the screenshot underneath; the caption’s job is to tell the reader why they should care about what they are looking at.

One small mechanical point: skip the full stop at the end. These are captions, not sentences, and a trailing dot makes a two-word phrase look oddly formal. Skip exclamation marks too, for different reasons.

How many screenshots should you actually make?

Three to five for most apps. Ten if your app genuinely has ten things worth saying, which is rarer than people think.

Apple accepts up to ten per display size and needs at least one. Google Play needs at least two phone screenshots to publish, and four before your listing can be promoted anywhere on the store. Those are floors, not recommendations.

A reasonable sequence for a five-screenshot set:

1. The main thing. The one feature that would make someone download the app.
2. The second-best thing, ideally one that addresses the obvious objection to the first.
3. A screen that shows the app is pleasant to use. This is the one that quietly does the design work.
4. Something that builds trust. Privacy, offline support, no account required, works on your other devices.
5. A soft close. Free to start, set up in a minute, whatever is true.

If you only have three, use the first, second and fourth. The trust screenshot earns its place more often than people expect, particularly for anything that touches money, health or personal data.

Device frames: yes, no, or it does not matter?

Neither store requires a device frame. Both are full of listings that use one and listings that do not, and both approaches have top-grossing apps behind them.

What a frame actually buys you is instant recognition. At thumbnail size, a framed screenshot reads as “an app” before the viewer has processed anything else. A bare screenshot can read as a photo, a web page, or a mistake, depending on what is on it.

Where frames hurt is space. The frame and its surrounding margin eat perhaps a third of the canvas, which is a third less room for your interface. If your app’s value is in dense information, like a chart or a calendar, a frameless screenshot that fills the image may simply show more of the thing worth seeing.

My rule of thumb: use a frame for the first two or three, where recognition matters most, and go frameless later in the set if a particular screen needs the space. If you are only trying out the look, the iPhone mockup generator will put a screenshot in a frame in a few seconds.

Consistency is most of what makes a set look professional

Take ten listings that look expensive and ten that look thrown together, and the difference is rarely the quality of the individual images. It is whether they look like they belong to each other.

One background treatment. One font. One caption position. One palette, ideally built from a colour already in your app icon so the whole listing hangs together. Changing the template between screenshot two and three reads as carelessness even when nobody can articulate why.

The strongest version of this is a background that runs continuously across the whole set, so the images line up when someone swipes through them. It is a small trick and it is immediately obvious when a listing has bothered.

Where you should vary things is the content, not the container. Different screen, different caption, same everything else.

What to leave off

This list is shorter than the list of things to include, but it costs more to get wrong, because some of these get the whole app rejected rather than merely underperforming.

• Star ratings you drew yourself. Both stores reject this. It also reads as desperate to anyone who has seen it before, which is everyone.
• Award badges and “featured” claims you were not actually given.
• Other platforms. “Also on Android” in an App Store listing is a reliable rejection, and the same applies in reverse.
• Long sentences. If the caption wraps to three lines it is too long, and the auto-sizing that keeps it inside the frame has already made it too small to read.
• Three ideas in one image. A screenshot that shows a chart, a list and a settings panel communicates nothing. One idea per image.
• Pure title cards. A set with no visible interface anywhere can be rejected, and it also fails the only job screenshots have, which is showing someone what using the app feels like.
• Features you have not shipped. Tempting when a release is close. Not worth the review cycle.

Google Play has one extra asset, and it is mandatory

If you are shipping to both stores, the one thing with no App Store equivalent is the feature graphic: a single landscape banner at exactly 1024 × 500 that sits at the top of your Play listing. Every listing needs one. Your app cannot be published without it.

The thing that catches people is that Play crops it. Different surfaces show different amounts, and some overlay their own elements near the middle. Anything that must be readable should sit well inside the image rather than near the edges. There is a dedicated Play feature graphic maker if you want that one on its own, and a Google Play screenshot generator preloaded with the Android sizes.

Localisation is worth more than another screenshot

If a meaningful share of your downloads come from somewhere that does not read English comfortably, translating your captions will usually do more than adding a sixth screenshot in English.

The layout, the frames and the actual screens stay exactly the same. Only the text changes, and both stores let you upload a separate set per locale. For most apps, Spanish, German and Brazilian Portuguese cover a lot of ground for very little effort.

One caution: machine translation is fine for a first pass but captions are the hardest kind of text to translate well, because they are short, idiomatic and doing a lot of work per word. Have a native speaker read them before a launch you care about.

A worked example

Say you have a habit tracker with three screens: a daily checklist, a weekly progress chart, and a streak counter.

A weak set captions these “Daily checklist”, “Progress chart” and “Streaks”. Accurate, and completely inert. Each one describes the screen to someone who is already looking at it.

A better set leads with the streak screen, not the checklist, because the streak is the emotional hook and the checklist is just a to-do list. Captions become “Keep your streak alive”, “See your week at a glance” and “Everything in one place”. Same three screens, same order of effort, substantially different first impression.

That reordering costs nothing and is usually the single biggest improvement available to a listing. The screens you are proudest of as a developer are often not the screens that sell the app.

Test it rather than arguing about it

Everything above is a reasonable default. None of it beats knowing what your own audience responds to, and both stores will tell you if you ask.

Apple has Product Page Optimization, which lets you run up to three alternative treatments against your current page and splits live traffic between them. Google Play has store listing experiments, which do much the same thing. Both report on installs rather than taps, which is the number that matters.

The mistake people make is changing everything at once. If you swap the template, the captions and the screenshot order together and conversions go up, you have learned that the new set is better and nothing else. Change one variable. The first screenshot is almost always the one worth testing first, because it carries the most weight.

Give it enough time to mean something. A test that runs for two days during a traffic spike tells you about those two days. Most apps need a couple of weeks before the numbers settle, and smaller apps need longer.

If your install numbers are too small for a meaningful test, borrow instead: look at the top ten apps in your category, note what their first screenshot says, and notice how few of them lead with a feature name.

Putting it together

Short captions about the reader. Strongest screen first. One consistent look across the set. Nothing on the images that is not true.

None of this requires a designer, and none of it requires more than an afternoon. What it does require is being willing to cut: the extra screenshot, the extra clause in the caption, the second idea you wanted to fit into one image. Store listings are one of the few places where doing less genuinely works better.

When you are ready to build the set, the App Store Screenshot Generator handles the mechanical half, including every required size for both stores. If you still need the icon to match, the App Icons Generator produces every iOS and Android size from a single image.