> ## Documentation Index
> Fetch the complete documentation index at: https://docs.geomaestros.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Competitive Gap Analysis
> Understand why AI mentions your brand - or why it doesn't - and get a clear action plan to close the gaps.
The Competitive Gap Analysis answers one straightforward question: **When someone asks AI about your industry, does AI mention you or not - and why?**
## The Core Concept: Wins vs. Losses
Every time a tracking run executes, your prompts are sent to the AI engines your plan includes (ChatGPT, Claude, Perplexity, Gemini, Google AI Overview and Google AI Mode). Each response is then evaluated:
* **Win** - you were the most-cited brand in the AI answer
* **Loss** - a competitor or other brand was the most-cited. You may still have been mentioned
The analysis then determines **why** you won or lost in each case. Note that a loss does not
mean you were invisible: the briefing splits losses into "appeared, lost" and "absent".
***
## What You See
### Win Rate (shown in your Executive Briefing)
A win is stricter than a mention. You "win" a prompt only when you were the most-cited brand in
that AI answer. Showing up but losing to a competitor is counted separately.
| Metric | What It Means |
| -------------------------- | -------------------------------------------------------------- |
| **Top answer** (green) | Prompts where you were the most-cited brand in the AI answer |
| **Appeared, lost** (amber) | Prompts where you showed up but a competitor was the top brand |
| **Absent** (red) | Prompts where you did not show up at all |
| **Win Rate** | Top answer divided by prompts analysed |
| **Appearance Rate** | (Top answer + Appeared, lost) divided by prompts analysed |
**Example:** Out of 14 analysed prompts you were the top answer on 6, appeared but lost on 5, and were absent on 3. That is a 42.9% win rate and a 78.6% appearance rate.
These totals live in the Executive Briefing. The Gap Analysis page itself shows the per-prompt
breakdown, why you won or lost each one, plus the cost and runtime of the analysis.
### Why You Win
When your brand IS mentioned, the analysis identifies what is working in your favor. These are your competitive advantages - the things you should continue doing. Examples include:
* Strong positioning as the recognized leader in your space
* High perceived relevance and clarity of what you offer
* Brand name that directly matches common search queries
### Key Gaps to Address
When your brand is NOT mentioned, the analysis pinpoints what is missing. These are your action items. Examples include:
* No dedicated page explicitly addressing a specific query topic
* Missing third-party discussions and endorsements
* Lack of structured content that AI can easily retrieve
The Gap Analysis complements your Share of Voice score. SOV tells you **where you stand**. Gap Analysis tells you **what to do about it**.
***
## How It Works
### AI responses are collected
Every tracking run sends your prompts to multiple AI engines and captures the full response, including which brands were mentioned, how many times, and which websites were cited.
### The most valuable responses are selected for deep analysis
The system prioritizes responses where competitors beat you, followed by emerging market opportunities, and then your wins. Each run covers as many distinct prompts as it can, around 36, so a large prompt set builds up coverage over successive runs rather than all at once.
### Each response is analyzed for patterns
For wins, the system identifies your strengths and what you should keep doing. For losses, it identifies what the winning competitor did differently and what specific improvements you should make.
### Results are deduplicated
The same prompt may be analyzed across multiple tracking runs. Only the latest analysis per prompt is kept, so you always see the most current competitive picture without inflated numbers.
### Findings appear in your briefing
The top winning factors and top gaps are surfaced in your Executive Briefing. You can also view the full Gap Analysis for complete detail.
***
## How Wins and Losses Are Determined
A "win" or "loss" is determined by who got the most mentions or citations in each AI response:
* **Mentions Win** - Your brand name appeared more times than any competitor in the AI response
* **Citations Win** - The AI cited your website more than any competitor's website in its sources
Coverage increases with each tracking run. If you recently added new prompts, they will be included in the next analysis cycle. Over time, every prompt in your workspace gets analyzed.
***
## How to Improve Your Win Rate
1. **Update your Google Business Profile** - Complete all fields, add photos, and respond to reviews
2. **Add FAQ pages** that directly answer your tracked prompts
3. **Include statistics and data points** on your website such as room counts, ratings, and distances to landmarks
4. **Get cited by authoritative sources** - Tourism boards, review sites, and industry publications
5. **Create content with quotes and testimonials** that AI can reference
6. **Use proper structured data markup** so AI can easily extract information about your brand
7. **Build brand mentions across the web** - Forums, social media, and press coverage
8. **Publish original research or data** that AI models will pick up over time
9. **Maintain consistent content updates** - Freshness signals matter to AI engines
***
## Understanding Confidence Levels
Each analysis comes with a confidence score from 0 to 100%. Higher confidence means the system is more certain about why you won or lost.
* **80%+** - Very reliable findings
* **60-80%** - Solid directional guidance
* **Below 60%** - Treat as directional hints rather than certainties
***
## How Often Is It Updated?
* Gap analysis runs **automatically after each tracking run**
* The briefing always shows the **latest result per prompt**
* Running more tracking runs means more prompts get analyzed and better overall coverage
* Results are cached for 7 days to avoid redundant analysis of the same patterns
Not all prompts are analyzed in every single run. The system prioritizes the most valuable analyses first and ensures every prompt gets covered over successive runs. Check back after a few tracking cycles for complete coverage.