Screenshots are a sequence, not a gallery
The most common way to think about store screenshots is as a set of pictures of your app. That framing produces a feature list: here is the home screen, here is settings, here is the profile page.
The better framing is a sequence with a job. Someone lands on your listing knowing nothing, spends a second or two, and either installs or leaves. Each frame has a distinct role in moving them along, and frames that do not have a role should not be there.
Two facts shape everything below. The first two frames do most of the work — in search results, those are often the only ones shown. And view rates fall sharply after frame three, so a great frame in position six is a wasted frame.
The five-frame structure
This is the order to default to. It works because each frame answers the question the previous one raises.
Frame 1 — The hook
The job is to earn the tap, and it has under a second to do it. This is the frame that appears as your thumbnail in search results, so it needs your strongest headline, your highest contrast, and the clearest single idea in the set.
It should answer "what is this and why do I care" without requiring frame two. If someone sees only this image, they should still understand your app.
Frame 2 — The one thing you do differently
Your most differentiated feature, told well. Not your longest feature list — the single thing a competitor cannot claim. This is where someone decides whether you are meaningfully different from the four other apps they are comparing.
Frame 3 — Deepen the story
Add a feature, a benefit or an emotion that builds on the first two without repeating them. The test for this frame is whether removing it would lose anything. If not, cut it and ship four frames.
Frame 4 — Proof
Anyone still looking has decided you might be worth installing and is now looking for a reason to trust you. This is where real numbers, genuine ratings and honest social proof belong.
Real ones only. Invented ratings, unearned "Editor's Choice" badges, and press logos from publications that never covered you are grounds for rejection, and they are transparent to anyone paying attention. If you have no proof yet, use this frame for a concrete benefit instead — an empty claim is worse than no claim.
Frame 5 — The close
This frame has the lowest view rate in the set, so it should be memorable rather than informative. An emotional image, a clear statement of the outcome, or a strong typographic closer. Do not put critical information here — most people will never see it.
Putting trust signals in one place
Ratings, badges and social proof belong on one frame, not sprinkled across all five. A star rating on every frame reads as insecurity. Concentrated on the proof frame, it reads as evidence.
How many frames to ship
Apple allows up to ten per device size. Ship five good ones.
The temptation with ten slots is to fill them, but every weak frame dilutes the set, and frames past the third are seen by a small minority. The only reason to go beyond five is if you genuinely have five strong, distinct things to say — and most apps do not, which is fine.
The rule that matters more than the count: never save your best material for later. Frame order should be strongest first, not a build-up.
The mistakes that cost installs
Leading with the home screen
A screenshot of your dashboard tells someone what your app looks like, not why they should install it. Lead with the outcome, then show the interface.
Frames that could belong to any app
If you could swap your app's name into a competitor's screenshots and nothing would look wrong, neither set is doing its job. Specificity is the entire game.
Raw, unframed screenshots
Uploading bare captures with no headline, no context and no design is the most common listing mistake and the easiest to fix. The screen alone does not explain itself to someone who has never used your app.
A set that does not look like a set
Five frames with different backgrounds, different type and a device that changes size between them read as five unrelated images. Consistency across the set is what makes it feel designed. The design guide covers the specifics.
Text too small to read at thumbnail size
Everything you put on a frame gets seen at roughly a tenth of its size first. If a headline cannot be read at that scale, it is not communicating — it is decoration.
Letting the set go stale
Screenshots showing an interface you shipped two redesigns ago actively mislead. People install expecting what they saw, find something else, and churn.
When a redesign is worth it
Redesigning a listing that already converts, with nothing else changed, is a gamble. These are the cases where it reliably pays:
- The screenshots no longer match the app. Non-negotiable — this costs you installs and retention simultaneously.
- The listing predates a rebrand. A store page in your old colours undermines everything else you have changed.
- Your biggest feature is missing. If the thing people love most is not in the first three frames, you are hiding your best argument.
- Conversion is below your category norm. Check App Store Connect's product page conversion rate. If it is materially below what is typical for your category, the visuals are usually the cheapest thing to change.
- The set was never designed. If it is raw captures from launch day, almost any deliberate set will beat it.
If none of these apply and your numbers are healthy, spend the effort elsewhere.
Testing rather than guessing
Opinions about screenshots are cheap and frequently wrong. Apple's Product Page Optimization lets you run up to three variants against your current listing and reports which converts better on real traffic.
Test one variable at a time. Changing the headline, the layout and the colour together tells you the new set won, but not why — so you learn nothing you can apply next time. Start with frame one, since it carries the most weight, and give each test enough traffic to mean something. Details in the A/B testing guide.
Launching and localising
Before you submit
Check the set at thumbnail size, confirm the frames reflect the build you are actually shipping, and make sure every claim on them is true. Verify dimensions against the current size requirements — a rejected upload on submission day is an avoidable delay.
Localisation is not translation
If you localise your listing text but leave English screenshots, the page reads as a half-finished port, and the screenshots are the part people actually look at.
The practical catch: translated copy is usually longer. German and Finnish headlines commonly run about 30% longer than English; Japanese and Chinese run shorter but need different line-breaking. A layout built tight around a seven-word English headline will break when that becomes eleven words. Design the frame with room, then translate — not the reverse.
Prioritise by revenue, not by language count. Two markets done properly beat eight done mechanically.
Frequently asked questions
What order should my screenshots be in?
Strongest first. Hook, differentiator, supporting feature, proof, close.
How many should I upload?
Five good ones. Ten slots exist, but view rates collapse after the third.
Where do ratings and badges go?
On one frame — the proof frame — and only if they are real.
Should the first screenshot show the app or a benefit?
Both: the benefit as the headline, the interface as the evidence beneath it. What it should not be is a bare screen with no explanation.
Get a set built this way
Nuvex builds the five-frame structure by default — hook, feature, deepen, proof, close — with the palette, type and device treatment locked across the set. About thirty seconds from your raw screens.