ASO for an app launch

Four weeks before submission

1. 1

   Build the query set

   Twenty to forty terms, scored for volume against how realistically a zero-history app ranks for them.

2. 2

   Split them across fields

   Highest-intent terms into the title and subtitle; the rest into the iOS keyword field, with no duplication.

3. 3

   Reserve a brand term

   A distinctive app name is what people search after seeing you elsewhere. Generic names are unfindable and untrackable.

4. 4

   Set up measurement first

   Store analytics, attribution and a tracked keyword list live before launch day, not after.

5. 5

   Prepare localization for launch markets

   Even three well-localized storefronts materially widen the indexed surface at zero marginal cost.

The first 90 days

Naming, and the one irreversible decision

Everything on a listing is editable, but the app name is the one field with brand consequences beyond the store, so it is the one worth arguing about before launch rather than after.

The pattern that works is brand plus category descriptor: it earns the brand search while telling the algorithm and the human what the app is. Pure brand names rank for nothing until you have marketing spend; pure keyword names look like spam and age badly.

Worked example: a productivity app six weeks before launch

A small team was four weeks from launching a focus timer and had not written a word of store metadata. The plan, as it usually is, was to ship the app and optimize the listing afterwards. That sequence wastes the single most valuable traffic period a new app gets, because launch attention arrives once and lands on whatever listing happens to be live.

Working backwards from the release date changed the order. Keyword research came first, and it changed the product name: the working title contained no searchable term, and adding one to the app name — the highest-weight field on both stores — was free before launch and awkward afterwards. Screenshots were drafted from the keyword set rather than from the design system, so the captions answered searches instead of describing panels.

The launch itself then had something to measure against. Because a baseline existed from day one — impressions, page views, conversion — the first post-launch iteration in week three was a real read rather than a guess made against no prior state, which is the position most new apps are in for their first quarter.

What the agent does in week one

1. 1

   Day 1 — research before naming

   Size the demand for the terms your app could plausibly own, and check the name against them while it is still cheap to change.

2. 2

   Day 2 — draft the full listing

   Title, subtitle, keyword field and description written from the keyword set rather than from the pitch deck.

3. 3

   Day 3 — brief the creative

   Screenshot captions derived from the searches you intend to win, so the frames answer intent instead of touring the UI.

4. 4

   Day 4 — set up measurement

   Define the baseline metrics and where they will be read, before there is any data to distort the choice.

5. 5

   Day 5 — schedule the first iteration

   Book the week-three review now, with the single change you will test already identified.