Gap analysis
Where AI perception diverges from your intended positioning.
Last updated: 2026-05-05
Gap analysis
Gap analysis compares what you claim about yourself with what AI models actually say about you. Every gap is a concrete content or positioning problem you can fix — and Pranas attaches a recommendation to each one.
How gaps are detected
- Pranas reads the brand positioning fields you filled in on the Brands page (problem solved, target customer, key differentiators, value proposition).
- It extracts the implied claimed attributes (e.g., "enterprise-grade security", "real-time collaboration", "no-code setup").
- It compares those against the AI-attributed attributes from the perception run using fuzzy matching (synonyms, related concepts, lemmatised forms — so "secure" matches "security" matches "safety").
- Any claimed attribute with zero or low AI attribution is flagged as a gap.
- Any AI-attributed attribute that contradicts your claims (e.g., you claim "affordable", AI says "expensive") is flagged as a misalignment (a stronger signal than a gap).
Reading a gap row
Each gap shows:
- Claimed attribute — what you said about yourself
- AI mentions — how often the AI actually echoed it (often 0)
- Closest AI attribute — the nearest thing the AI did say
- Severity — High / Medium / Low based on how central the attribute is to your stated positioning
- Recommendation — a generated content brief or action item
Gap → Perception bridge
Inside the Brand Perception page, a "Gap Perception Bridge" widget pairs each gap with the matching driver theme so you can see, for example, that your "enterprise security" claim has zero AI coverage and that your competitors dominate the security driver. That's a clear remediation target: ship security content and pursue citations on security-focused domains.
Gap types reference
| Finding type | What it means | Strength of signal | Typical fix |
|---|---|---|---|
| Gap | You claim an attribute, the AI gives it zero or low attribution | Moderate | Publish content that evidences the claim |
| Misalignment | The AI says the opposite of your claim | Strong | Reposition or correct the public record |
| Inverted gap | The AI attributes a strength you don't claim | Opportunity | Add it to your own marketing |
| New-product gap | Claim is real but training data hasn't caught up (first 4–8 weeks) | Low urgency | Wait and pursue citations |
| Stale positioning | The same gap persists for multiple quarters | Strong | Rewrite the positioning pages |
Common patterns
- New product, no gap fix yet — typical for the first 4–8 weeks after launch; AI training data hasn't caught up. Reduce the urgency for these.
- Stale positioning — gaps that persist for multiple quarters usually mean your positioning isn't reflected in your published content.
- Inverted gap — the AI attributes things to you that you don't claim. These are often strengths you're underselling in your own marketing.
Acting on gaps
The Action Center pulls high-severity gaps directly into its recommendation queue. Each recommendation includes:
- Suggested page to write or update
- Target keywords
- Citation domains to pursue
- Estimated impact
You can dismiss recommendations or convert them to Content Studio briefs.
Related articles
- Brand Perception overview
- Perception scoring methodology
- Action Center
- Brand positioning fields