How to get more app downloads

If you want to know how to get more app downloads, start with your store listing, not your budget: an icon that stands out, well-distributed keywords and screenshots that win people over in the first two images. Then grow your rating volume, gain visibility in browse with In-App Events, Custom Product Pages and promotional content, and open up new countries and languages.

The last step is the one that has changed the most: working on how your app is talked about outside the store, because more and more people discover apps by asking an AI before they ever open the App Store or Google Play. And you measure all of this by source and month by month, not by a single keyword ranking or week by week. It's the method I use every day with the apps I manage; here it is, step by step.

Where your downloads come from

Before you change anything, split your downloads by source. These are the ones I look at:

• Search: users type something and find you. Keywords rule here.
• Browse: they discover you while exploring the store: the Today tab, Games, Apps, charts or recommendations. Your icon, events and featuring rule here.
• Referrals and web: they arrive from a link on another site, a blog, social media or your own website.
• AI assistants: more and more people ask ChatGPT or another assistant "which app do you recommend for…" and reach the store with their mind already made up. The consoles don't break it out as its own source, but you still need to work on it.

You'll find the first three in App Store Connect and Google Play Console. Each source has its own lever. If you don't know which one is failing, you're flying blind.

1. Fix your listing before you pay for traffic

The most expensive mistake I see is spending on UA with a listing that doesn't convert. Every euro or dollar you spend on paid goes less far if your listing loses the people who land on it. The order is always the same: fix the listing first (icon, screenshots and metadata), then scale paid. If that's you and your campaigns are already running, an ASO audit will tell you what's holding back conversion before you keep paying.

Icon

It's the first thing every user sees, organic or paid, so it's the first thing I test. In Soccer Star 2018 World Legend, by Redvel Games, we swapped the icon during Ramadan for one featuring the player Salah and a Ramadan crescent moon, and downloads went up 30%. On the App Store you can test icons with Product Page Optimization, with up to 3 treatments against the original for up to 90 days (Apple); on Google Play, with store listing experiments, with up to 2 variants against your current listing (Play Console Help).

Keywords on the App Store and Google Play

Apple ranks your app in search based on text relevance (title, subtitle, keyword field and primary category) and user behavior: downloads, ratings and reviews, among other things (Apple). My rules for the App Store:

• The main keyword, the one that defines your category or that you simply have to rank for, always goes in the title.
• I never repeat keywords across the title, subtitle and keyword field. Apple also asks you not to repeat words already in your app name, subtitle or category.
• I never use competitor brand names in metadata or in the keyword field.

Google Play has no keyword field: the full description carries a lot of weight and leaves room for long-tail keywords. If you want to go deeper into each store, I cover it on my App Store ASO and Google Play ASO pages. These are the official limits:

To check yours without doing the math, use my ASO character counter: it shows what's left in each field, the bytes in your keyword field and which words you're repeating.

Screenshots and video

With no app preview, the first one to three screenshots appear in App Store search results, depending on their orientation (Apple). That's why the first two are the ones that convert: they have to show the most powerful thing about your app. In a social casino app, new screenshots lifted listing conversion by 3%.

There's one nuance I'm more and more sure about, and I shared it in Applift's 2026 app marketing predictions, where I was one of the 7 invited experts: with everyone using the same AI tools to get perfect copy, screenshots and visuals, perfection has lost its value and become generic. For many apps, that polished hyper-marketing style backfires, and sometimes simpler, more human, "homemade" screenshots and copy work better. Imperfect but honest.

2. Ratings and reviews

Ratings count twice: Apple lists them among its search factors, and they're what everyone checks before downloading.

When to ask for a rating

I always trigger the rating prompt after a positive moment: an achievement unlocked, reaching a certain level, winning something. Never right after install and, ideally, not on day one either. Doing it this way always increases review volume, and because the user has just had a good experience, the reviews tend to be good. This is what the platforms say:

• Apple: you must use its API to ask for ratings; Apple doesn't allow custom review prompts (guideline 5.6.1) or forcing users to rate (3.2.2) (App Review Guidelines). The system shows the prompt at most 3 times in 365 days and decides whether it appears, so don't trigger it from a button (RequestReviewAction). Apple recommends asking once people have genuinely engaged with the app, never on first launch or during onboarding (Human Interface Guidelines).
• Google: the In-App Review API has a time-bound quota that Google doesn't publish; you shouldn't trigger it from a button, ask questions beforehand such as "Do you like the app?", or modify the card (Android Developers). And its policy prohibits offering incentives for ratings (Play Console Help).

Replying to reviews and listening to complaints

I always reply to reviews in a friendly tone, trying to offer a fix and keep the user in the app. Both stores let you do it: in App Store Connect you can respond to each review, and Apple notifies the user, who can then update it (App Store Connect Help); on Google Play you can post one public reply per review (Play Console Help). And I always pass recurring complaints to the development team as a priority: a negative review that keeps coming back is a bug or a missing feature that's costing you downloads.

Crashes and ANRs

Another recurring mistake is investing in the listing while crashes, ANRs and bad reviews pile up. On Google Play, the bad behavior threshold is a 1.09% user-perceived crash rate and a 0.47% ANR rate; above that, your app may be less discoverable and show a warning on its listing (Android Developers). Neither Apple nor Google is going to promote a buggy app for free.

3. Win visibility in browse

This is where my work has changed the most in recent years, and I said so in Applift's predictions too: I'm giving more priority than ever to In-App Events, organic Custom Product Pages and anything that improves browse visibility.

• In-App Events (App Store): I launch at least one a month, plus key dates like Christmas, Easter, Thanksgiving, Halloween or the 4th of July. Each event can last up to 31 days and shows up in search, across the store tabs and on your product page (Apple). They work best for winning back lapsed users: in a Christmas campaign for a social casino app in the US, the event passed 200,000 impressions in one month and over 99% of interactions came from people who already had the app.
• Organic Custom Product Pages (App Store): you can publish up to 70 and assign keywords to them so they appear in the results for those searches instead of your default product page (Apple). In a social casino app we used them for very specific long-tail searches, such as Caesars-style slots or Las Vegas and Rome themes, with screenshots in that theme and copy that invites people to download. The results in conversion and downloads have been very good.
• Promotional content (Google Play): offers, time-limited events and major updates that appear across the store, in search results and on your listing. It's available for all games and for apps that meet Google's eligibility criteria (Play Console Help).

Running that calendar every month is exactly what I do as a fractional ASO manager.

4. Localize and open new countries

Another mistake I see a lot is skipping localization or picking the wrong countries, leaving entire markets unoptimized. When I joined CoinMarketCap, the app wasn't optimized in any country; we expanded it to global markets, going from tier 1 all the way down to tier 4 countries, to gain impressions.

AI helps a lot here, but only with the boring part: resizing, localization, optimizing keywords for 20 languages in seconds or predicting which assets are most likely to win before you test. I said it in Applift's predictions: it won't replace app marketing, it'll raise the bar for juniors. The value is in strategy, in understanding user psychology, reading the market and turning those insights into real growth.

And never publish without reviewing. I launched my app Ferro with metadata in 30 languages, and machine translation turned "lift" into an elevator and "PR" (personal record) into public relations. Since then I review every language, especially in priority markets. If you want to get it right from the start, I explain how on my ASO localization page.

5. Beyond the store: from Search to Discovery Optimization

People don't browse the stores the way they used to. In Applift's predictions I put it like this: we're moving from "Search Optimization" to "Discovery Optimization". Many users will get app recommendations from AI assistants before they even open the store, so your metadata has to make sense to LLMs as well.

Get your app talked about on the open web

I try to get the app mentioned on blogs, on its own website and anywhere on the open web: the more places, the better. Algorithms increasingly use those external signals to understand an app's context and authority. It's no longer just the keyword in your title: it's the consensus the internet builds about your app.

Metadata an AI can understand

If an AI assistant can't clearly tell what your app offers, you basically don't exist. Apple already shows app tags in search results that are generated with LLMs from your App Store Connect metadata (Apple). Write your title, subtitle and description so that anyone, person or model, understands in one sentence what your app does and who it's for.

Editorial featuring

You can ask for it, but it's hard to get; if the app is good and does something new, it's possible.

• Apple: you request it in App Store Connect with a featuring nomination, at least 3 weeks ahead and, for wider consideration, up to 3 months before (App Store Connect Help, Apple). If you're preparing a launch, put it in the calendar: it's part of what I do in an app launch.
• Google Play: it doesn't accept self-nominations (Play Console Help). The official route is to request featuring when you create a promotional content event, if your app or game meets Google's premium growth tools criteria, at least 14 days in advance and with no guarantee you'll be featured (Play Console Help).

Alternative app stores

I've tried Samsung Galaxy Store, Amazon Appstore and Huawei AppGallery, and the one that has worked best for me is Huawei AppGallery. Amazon shut down its Appstore for Android devices on August 20, 2025, and it now only lives on Fire TV and Fire tablets (Amazon).

6. Measure what matters

Stop reporting keyword rankings as your main KPI. I said it in Applift's predictions and I apply it every month: obsess over conversion by source and LTV, and know who your users are, where they come from and how much value they bring over time. To do that, I use Sensor Tower, Looker, App Store Connect, AppsFlyer and Google Play Console every day.

Two more rules:

• Change things in blocks. Another classic mistake is changing everything at once (icon, screenshots and keywords) and never knowing what worked. I change one block at a time and measure each change.
• Don't judge after a week. In organic, a week is nothing. The App Store needs at least a month and a half; Google Play, at least three months.

The mistakes I keep seeing

I've seen them in my own apps, in clients' apps and in teams I've worked with. Run through them before you touch anything:

• Are you paying for UA with a listing that doesn't convert yet?
• Have you changed several things at once and can't tell which one worked?
• Are you drawing conclusions from a week of data?
• Do you have entire countries or languages left unoptimized?
• Do you have unresolved crashes, ANRs or bad reviews while you invest in your listing?

If you make games, most of this applies, but the genre has its own levers. I cover them in my guide on how to get more downloads for your mobile game and on my mobile game ASO page. And if you're an indie developer short on time, start with the icon, then keywords, then screenshots; my ASO for indie developers page explains how I work with small teams.