The GEO Readiness Score is a 100-point framework we use to assess how ready a website is to be found, understood and cited by AI assistants such as ChatGPT, Perplexity, Gemini and Google’s AI Overviews. It scores seven areas: crawler access, rendering, entity clarity, answerable content, original information, off-site presence and measurement. Anyone can apply it to their own site in an afternoon. The weights are Brilad’s judgement based on how these systems are documented to work, not a formula published by any search or AI company.
Why a score helps
GEO advice is usually a list of tips with no order. A score forces two useful things: it shows which area is weakest, and it gives a baseline to measure against three and six months later. It also stops teams from spending weeks on llms.txt or schema while their firewall blocks the crawlers that matter.
The seven areas and their weights
GEO Readiness Score: points per area
Total 100 points. Weights reflect how early each area sits in the chain from crawling to citation.
- 1. Crawler access20
- 2. Rendering and readability15
- 3. Entity clarity15
- 4. Answerable content20
- 5. Original information10
- 6. Off-site presence10
- 7. Measurement10
Source: Brilad GEO Readiness Score, version 1, September 2026.
Crawler access and answerable content carry the most weight because failure there stops everything else. A site with perfect content scores nothing in practice if OAI-SearchBot receives a 403.
1. Crawler access (20 points)
| Check | Points |
|---|---|
| robots.txt allows Googlebot, Bingbot, OAI-SearchBot, Claude-SearchBot and PerplexityBot on key pages | 6 |
| CDN and firewall return status 200 to those crawlers, with no challenge pages | 6 |
| Server logs show visits from AI search crawlers in the last 30 days | 4 |
| Key pages return 200 with no noindex, login wall or redirect loop | 4 |
Test with our AI crawler access checker.
2. Rendering and machine readability (15 points)
| Check | Points |
|---|---|
| Main text of key templates is in the HTML before JavaScript runs | 6 |
| One H1 and a logical H2 and H3 outline on key pages | 3 |
| Prices, specifications and comparisons in HTML text or tables, not images | 3 |
| Tabs, accordions and counters contain their real content in the HTML | 3 |
3. Entity clarity (15 points)
| Check | Points |
|---|---|
| Company name, address and phone identical on the website, Google Business Profile, LinkedIn and directories | 5 |
| One Organization schema node with sameAs links to official profiles | 4 |
| An About page naming the people behind the business, with Person schema | 3 |
| Searching the exact brand name shows no confusion with other companies | 3 |
4. Answerable content (20 points)
| Check | Points |
|---|---|
| Each core service or product page answers what it is, who it is for and what it costs or what drives the cost | 6 |
| Sections open with a direct answer to the question in their heading | 5 |
| Pages exist for the main buyer questions, including comparisons and problems | 5 |
| Passages name their subject and make sense on their own | 4 |
5. Original information (10 points)
| Check | Points |
|---|---|
| At least three pages contain data, results, tests or tools that competitors do not have | 5 |
| Statistics name their source and date | 3 |
| Named authors with real experience on articles | 2 |
6. Off-site presence (10 points)
| Check | Points |
|---|---|
| Listed with current details on the review and directory sites that assistants cite for your category | 4 |
| Genuine reviews from customers on at least one independent platform | 3 |
| Mentioned by name in industry publications, events or partner sites in the last year | 3 |
7. Measurement (10 points)
| Check | Points |
|---|---|
| A fixed set of buyer questions tested monthly across assistants | 4 |
| Analytics separates visits from AI assistants | 3 |
| Search Console connected and reviewed monthly | 3 |
The method for the first two is in how to track brand mentions for free.
Reading your score
| Score | What it usually means | Focus next |
|---|---|---|
| 0 to 39 | AI systems probably cannot reach or understand the site | Areas 1 and 2 before anything else |
| 40 to 64 | Reachable, but rarely the best answer | Areas 3 and 4 |
| 65 to 84 | A solid base; citations depend on distinctiveness | Areas 5 and 6 |
| 85 to 100 | Well prepared; gains come from consistency over time | Area 7 and steady publishing |
How to use it well
- Score areas 1 and 2 separately for each template type: home, service or product, article and listing. One broken template can hide hundreds of pages.
- Give partial points only when a check is partly met on the pages that matter most.
- Re-score every quarter and keep the evidence: screenshots, log extracts, test results.
- Treat the score as a way to order work. It does not predict citations, and no score can.
Frequently asked questions
Is this an official score?
No. It is Brilad’s framework. Google, OpenAI and others do not publish a readiness score. The checks come from their documentation and from published research, and the weights are our judgement.
Why are llms.txt and FAQ schema not included?
No major AI provider has said it uses llms.txt to choose sources, and FAQ rich results in Google are limited to government and health sites. Neither is a reliable driver of AI visibility, so neither earns points.
What score should we aim for?
Above 65 before investing heavily in content, since below that the basics limit results.
Sources
- Google Search Central, “AI features and your website”
- OpenAI, “Overview of OpenAI crawlers” and “Publishers and Developers FAQ”
- Aggarwal et al., “GEO: Generative Engine Optimization”, 2023
- Seer Interactive, “What drives brand mentions in AI answers?”