Is instinct-based keyword selection always wrong?

No — instinct is useful for generating the candidate list. The problem is using instinct for the final selection, without demand and difficulty data to validate the choices. Instinct says 'habit tracker is relevant'; data says whether 'habit tracker' or 'daily habits app' has better demand and accessibility.

At what point does keyword research pay off vs. just guessing?

When you have an app generating some organic installs and you want to grow them systematically. For a pre-launch app or a very early-stage product, instinct plus basic metadata hygiene is a reasonable starting point. Once you're investing in ASO updates to drive growth, data-driven research pays for itself.

Can't I just check my keyword performance in App Store Connect?

You can see which terms led to impressions and installs after the fact — but not the demand or difficulty of terms you're not yet targeting. App Store Connect analytics tells you how well your current keyword set is working; keyword research tells you which set to try next.

Compare

ASOGrade vs. Guessing on Instinct

Moving from gut feel to grounded keyword decisions

You're picking keywords on instinct.

Instinct is fine for a first guess and expensive for a metadata slot. Every field you fill without a demand number is a slot you cannot measure.

Most developers pick their first App Store keywords on instinct. You know your app, you know what it does, and you write the subtitle and keyword field based on how you'd describe it. This is a reasonable starting point — you're not making things up, you're drawing on genuine product knowledge.

The problem is that instinct doesn't have access to two pieces of information that matter: which of your instinctive terms actually have search demand in the App Store, and which of them have ranking sets you can realistically break into. Without those numbers, you're optimizing on assumptions that may or may not match real user search behaviour.

What this approach does well

Instinct-based keyword selection draws on your understanding of your app's category, your competitors' names, and the language you see in similar apps' subtitles. These are valid inputs.

It is genuinely faster for a first pass than running a full research workflow. For very early-stage apps where the entire product is still being validated, spending an afternoon on keyword research before you know if anyone wants the app at all is a poor use of time.

The honest cost of instinct-only selection: you're choosing without knowing whether a term has 5 searches a day or 500, and without knowing whether the top-ranking apps for that term are beatable by an app with your current rating count.

Where it falls short

Instinct reliably picks category-correct keywords — terms that describe what the app does. What it misses is the distinction between keywords that describe the app and keywords that describe the search. These are often different. Users search for problems and outcomes ('sleep better', 'stop snooze') not always for product categories ('sleep timer', 'alarm app').

Instinct also has no signal on difficulty. You might choose a perfectly relevant keyword that is dominated by apps with 50,000 ratings and a three-year ranking history. That term is in your metadata, you're indexed for it, and you rank 45th. The traffic value is near zero.

Over time, the instinct approach generates a keyword set that is semantically reasonable but not optimised — some terms are generating installs, most aren't, and you have no data to tell them apart.

Where ASOGrade fits

ASOGrade replaces the guess on demand with a number: the Apple Search Ads popularity score for each keyword in each storefront. Before writing a single character of metadata, you can see which of your instinctive candidates actually have users searching for them.

It also replaces the guess on accessibility with the difficulty score. You can see, before committing to a keyword, whether the apps currently ranking for it are beatable at your current stage — or whether you should find a lower-difficulty synonym first.

The workflow is: generate your instinctive candidates (instinct is still useful for ideation), score them, cut the ones with no demand or impossible difficulty, and write metadata against what's left. The instinct drives the candidate list; the data drives the final selection.

Frequently asked questions

Is instinct-based keyword selection always wrong?: No — instinct is useful for generating the candidate list. The problem is using instinct for the final selection, without demand and difficulty data to validate the choices. Instinct says 'habit tracker is relevant'; data says whether 'habit tracker' or 'daily habits app' has better demand and accessibility.
At what point does keyword research pay off vs. just guessing?: When you have an app generating some organic installs and you want to grow them systematically. For a pre-launch app or a very early-stage product, instinct plus basic metadata hygiene is a reasonable starting point. Once you're investing in ASO updates to drive growth, data-driven research pays for itself.
Can't I just check my keyword performance in App Store Connect?: You can see which terms led to impressions and installs after the fact — but not the demand or difficulty of terms you're not yet targeting. App Store Connect analytics tells you how well your current keyword set is working; keyword research tells you which set to try next.

Other comparisons

Try the research pass on its own

Paste 100 keyword ideas and read popularity, difficulty, and competing app count across 109 storefronts — in seconds.

Get started