What app marketing decides
Three decisions sit inside app marketing, and most teams make them in the wrong order. The first is what the app is for, in the store's terms: which searches it should appear in, which category it belongs to, and which single promise the listing makes. The second is how the listing converts: the title, subtitle, screenshots, previews, ratings and localisation that turn a store visitor into an install. The third is how much reach to buy, on which networks, optimised to which event, at what cost. The first two together are the work known as ASOImproving an app's visibility in app store search and the share of store visitors who install it., app store optimisation.
The order matters because the third decision is judged through the second. A paid campaign cannot outperform the listing it sends people to; a weak listing wastes every click. Apple and Google both report store conversion by source in their consoles, which makes the dependency visible: when paid traffic converts far below organic, the problem is usually the match between the ad's promise and the listing, not the campaign.
What app marketing does not decide is what happens after the install. Activation, retention messaging and measurement belong to the messaging and measurement sections, and the wider mobile marketing hub sets the whole loop out. This section stops at the install and explains how to make each one count.
Store presence, paid installs, launch
The three ways to get installs have different costs, speeds and ceilings. Most apps need all three at some point, in different proportions as they mature.
| Lever | What it does | What it costs | Where it is weak |
|---|---|---|---|
| Store presence (organic) | Ranks the app for store searches and browse placements; converts visitors on the listing | Research and creative effort; slow to build, cheap to keep | Bounded by how many people search for the category; slow to change |
| Paid install campaigns | Buys reach on ad networks whose formats open the store; scales with budget | Media spend, creative production, measurement setup | Expensive where competition is high; quality falls as volume is pushed |
| Launch and editorial | Concentrates attention in a short window through store features, press, creators and existing audiences | Preparation time, relationships, a launch-worthy release | One-off; a feature cannot be bought and does not repeat on demand |
A launch is the moment the three overlap. Store features are editorial decisions that the stores document as being made on quality, design and newsworthiness, not on request, so the realistic launch plan is a listing that is ready, a paid campaign that is tested and paused until launch day, and the business's existing channels (email, web, social) pointed at the store at once. The spike a launch produces is useful mostly for what it teaches about conversion and early retention at volume. The launching a new app guide plans that window, from readiness before submission to the first month after release.
The core guides
- App store optimisation — how store search and browse surface apps, keyword research for stores rather than search engines, what the title, subtitle and description each do, screenshots and previews as the conversion asset, ratings and replies, and localisation and the stores' own listing tests.
- App install campaigns — the campaign types each network offers for apps, optimising to installs against in-app events, creative that works in app ad formats, budgets, bidding and pacing, the store listing as the landing page, and install fraud.
The core guides are written to be read together. The optimisation guide assumes the listing is the destination for paid traffic as well as organic; the campaigns guide assumes the listing has been fixed first. The illustrative TikTok app-install case then walks through one invented campaign from brief to decision, with every figure made up, and is the worked example neither guide repeats.
Further guides
Further guides cover the decision before the app exists, the routes into it, the work around the listing and who does that work:
- App or mobile website — deciding between building an app and improving the mobile website: what an app adds, what it costs, the jobs that need one, and the hybrid route.
- Web-to-app journeys — moving website visitors and customers into the app: where app prompts belong, links to the matching screen, signed-in continuity, and measuring the switch.
- Building custom store listings for campaigns — a tutorial for listings built for campaigns: choosing themes, matching screenshots to the ad, linking each campaign, and reading conversion per page.
- Setting up app rating prompts and review replies — a tutorial for the rating routine: when to ask, who never to filter out, triage, replies that help readers, and fixes by version.
- Launching a new app — readiness before submission, building demand before release, launch week across owned channels, and the first month.
- Hiring and evaluating an app marketing agency — what a provider is hired to do, the accounts that stay in the business's name, questions and warning signs for a proposal, reports that tie back to the store consoles and the measurement partner, and handing over.
The install-to-retention loop
An install is the cheapest event an app produces and the one that says least. The pirate-metrics framing (AARRR) puts acquisition first only so that activation, retention and revenue can be read against it. In practice the loop runs like this: store and paid sources produce installs; a share of installs reach the activation event; a share of those are retained; retained users produce value and, if asked well, ratings and referrals; the ratings lift listing conversion (CVRShare of visits or clicks that convert.) and the value funds the next round of paid acquisition at a cost per acquisition (CPACost of each conversion.) the business can carry.
The practical consequence is that app marketing has to be judged on something deeper than installs. Cost per install rewards whoever buys the cheapest users; cost per activated user, or per retained user at a set day, rewards whoever buys the right ones. Both core guides in this section treat the deeper event as the real target, as does the launch guide, and the measurement section explains how to produce it.
Pellbrook Languages reads two sources by activation
Pellbrook Languages, an invented language-learning app, runs paid installs on two networks for a month with £4,000 on each. Network A delivers 2,600 installs at £1.54 each; Network B delivers 1,500 at £2.67. On install cost alone, A wins by a wide margin. The analytics show that 19% of A's installs complete a first lesson against 41% of B's, and seven-day retention is 8% for A and 23% for B. Cost per activated user is £8.10 on A and £6.50 on B; cost per user retained at day seven is £19.23 on A and £11.59 on B. Pellbrook moves most of the next month's budget to Network B, and asks Network A to optimise to the lesson-completion event rather than to installs before it gets a second chance.
Common app marketing problems
Most app marketing problems show up as one of a few symptoms. The table maps each to the guide that diagnoses it.
| Symptom | Usual cause | Guide |
|---|---|---|
| Few store impressions for category searches | Title and subtitle not built around searched terms; keyword field wasted on brand | App store optimisation |
| Store visitors who do not install | First screenshots do not show the promise; weak or stale ratings; no localisation | App store optimisation |
| Paid installs converting far below organic | Ad promise and listing promise differ; wrong product page as destination | App install campaigns, App store optimisation, Custom store listings for campaigns |
| Cheap installs that never activate | Campaign optimised to installs on a network that finds low-intent users; fraud | App install campaigns |
| Low click-through (CTRShare of impressions that result in a click.) on app ads | Creative that describes rather than shows; no hook; wrong format for the placement | App install campaigns |
| Installs fall after a launch spike | Launch traffic was borrowed attention, not a repeatable source | App store optimisation, App install campaigns, Launching a new app; diagnosis: organic app installs falling |
| Nobody can say which installs were worth buying | No activation event or measurement partner in place before spend | Mobile measurement |
| Rating stale or falling; complaints left unanswered | Prompt asked at first launch or not at all; replies only to praise; bugs never reach product with the app version | App rating prompts and review replies |
| Website visitors shown the app and leaving | Prompts shown to every visitor or as a full-screen overlay on arrival; the switch opens the home screen | Web-to-app journeys |
| An app proposed to fix a mobile site that fails on phones | The website's problem moved behind an install step; the hybrid route never scored | App or mobile website |
| An agency's or freelancer's reports look good but the business cannot check them or keep the accounts | Store, measurement partner or network accounts opened in the provider's name; installs reported without activation | Hiring and evaluating an app marketing agency |
When several of these symptoms appear together, the app marketing audit scorecard scores the listing, campaigns, first session, messaging and measurement side by side and ranks the fixes.
