Keyword Scoring

When you run a keyword analysis in Apptonomy, the platform examines your app and up to ten competitors to produce a prioritized list of keyword opportunities — the terms real people are searching for where your app has the best chance of climbing the rankings. Here’s how that process works and what each score means.

From Hundreds of Keywords to a Focused List

Think of the analysis as a funnel. It starts wide — collecting every keyword that could be relevant — and progressively filters down to the ones that actually matter.

1. Extraction

The engine scans the title, subtitle, description, and screenshot text of your app and each competitor. Keywords are identified in two ways: directly from the text itself, and through AI analysis that infers terms a human optimizer would associate with the app — even if those exact words don’t appear in the listing. This works across Latin, Chinese, Japanese, Korean, Thai, and Arabic scripts.

2. Deduplication

Duplicate keywords are merged. When the same keyword appears across multiple apps, those overlaps are tracked so you can see which competitors share each term.

3. Data Enrichment

Each keyword is scored across four dimensions using real-world data:

• Volume — how many people search for this term each month
• Difficulty — how many competitors are fighting for it
• Rank — where your app currently appears in search results
• Trend — whether search interest is rising, stable, or declining

The engine also flags branded keywords (like “instagram” or “adobe photoshop”) so you know which terms belong to another company.

During this stage, the engine collects suggested keywords — related terms recommended by ad platforms that weren’t in the original extraction. Suggestions that appear frequently across multiple keywords are treated as stronger signals.

4. Quality Filtering

Suggestions are filtered by an AI quality check that removes generic or non-actionable terms — words like “get,” “world,” or “experience” that rarely represent a real search intent. The best surviving suggestions are enriched with the same metrics and added to the pool.

5. Volume Filtering

Keywords with little or no search volume are removed. This eliminates terms that might be relevant in theory but have no meaningful search audience in practice.

6. Relevance Scoring

An AI model evaluates each remaining keyword in the context of your app’s description and assigns a relevance score from 1 to 100. Feature-specific terms, category descriptors, and user-intent phrases score high. Generic filler words and unrelated terms score low. Keywords below the relevance threshold are removed to ensure the final list only contains terms that genuinely connect to what your app does.

The Opportunity Score

Every keyword that survives filtering receives an Opportunity Score from 0 to 100. This is the single most important number for deciding where to focus your efforts. It balances three factors:

A high Opportunity Score means the keyword has a healthy search audience, manageable competition, and ideally a rising trend. It’s a term where optimizing your listing is most likely to pay off.

A low Opportunity Score means the keyword is either too obscure, too competitive, or losing momentum.

Reading the Difficulty Score

Difficulty runs from 0 to 100, and lower is better — a low score means you can realistically compete for the term. In the keyword tables the score is color-coded into three bands so you can scan a list at a glance:

The score reflects how strong the apps already ranking for the term are — their ratings, review activity, age, and update cadence — combined with how hard advertisers bid on it. A term held by a few dominant apps scores higher than one spread across weaker listings.

Storefront and Store Calibration

Volume and difficulty are not universal numbers — they are calculated for the exact storefront (country) and store (App Store or Google Play) you’re analyzing. The badges next to the page title tell you which market a result belongs to.

• Storefront: Search demand and the set of competing apps differ by country, so the same term can be low difficulty in one storefront and high in another. Analyze each market you care about separately.
• Store: Difficulty is computed from the actual top search results on that store, and volume comes from Google Ads calibrated against Apple Search Ads. A keyword’s numbers on the App Store won’t necessarily match its numbers on Google Play.

Your Results

The final output splits keywords into two lists:

• Existing keywords — terms your app already ranks for, sorted by opportunity. These show where to protect or improve your current position.
• Suggested keywords — terms your app does not rank for, sorted by opportunity. These are new opportunities worth targeting.

Each keyword includes its volume, difficulty, opportunity score, relevance score, and a breakdown of which competitors rank for it.

The Overall Keyword Score

Beyond individual keywords, you’ll see a single summary score from 0 to 100 that reflects your app’s overall keyword health. It combines:

• Ranking quality (40%) — What share of your existing keywords rank in the top 10? Top 50? More high rankings mean a stronger foundation.
• Opportunity potential (60%) — How much untapped opportunity remains? High opportunity scores among suggested keywords signal significant room to grow.
• Coverage bonus — A small bonus for the breadth of keywords you already rank for, reflecting overall visibility.

Where the Data Comes From

Related

• Keywords & Clusters
• Keyword Engine
• Glossary