Your Match Protocol Score™ is a 0 to 100 number that shows how well AI finds, matches, recommends, verifies and hands off your business. Read it in three passes: tier first, then stage, then fix the weakest stage that carries the most weight.
What does the Match Protocol Score™ measure?
The Match Protocol Score™ measures how ChatGPT, Google AI Overviews, Perplexity and Gemini treat your business when a buyer asks a question that never names you.
Each business gets up to 32 blind buyer conversations: up to 8 questions across 4 engines. The score is built under methodology version 2.0. You can read the full approach on the Match Protocol Score™ page.
What are the five stages?
The score is made of five stages, and each one answers a single question about your business.
- Found (20%): Can AI read you?
- Matched (20%): Does AI name you for the right buyer questions?
- Recommended (35%): Does AI recommend you?
- Trusted (15%): Can AI verify you?
- Chosen (10%): Is your site built to convert the human?
Recommended carries the most weight because a recommendation is the moment AI sends a buyer your way. Chosen carries the least because it only matters once the buyer shows up.
What do the four tiers mean?
The tiers tell you, in one word, how AI treats you today.
Ghosted (0 to 49): AI rarely or never names you. Friend Zone (50 to 69): AI knows you but seldom picks you. Second Date (70 to 84): AI recommends you regularly and you are close. The One (85 to 100): AI names you for the buyers who matter.
In the Match Protocol™ Benchmark 2026, 60% of 100 small businesses were Ghosted and 39% were in the Friend Zone. Just 1% reached Second Date and none reached The One.
Why can't a business that AI never names score above 49?
Because the score is capped by the AI Recommendation Rate.
A clean website and a long list of directory listings do not matter if AI never names you in a buyer conversation. The cap keeps the score honest. It forces the number to reflect what the engines do, not what your site looks like.
Which stage should you fix first?
Fix the earliest stage that is weak, then the heaviest one.
Order matters because each stage depends on the one before it. If Found is low, start with Machine-Readable Architecture, the M layer of Match Protocol™. Nothing else works if AI cannot read you. If Found is solid and Matched is low, go to Algorithmic Alignment, the A layer, and map the buyer questions you are missing.
If Found and Matched are fine but Recommended is low, work on Targeted Extraction, the T layer. Lead with answers and add FAQ content. This is where most owners land. The benchmark stage averages were Found 80, Trusted 87 and Recommended 15.
If Trusted is low, build Citation & Corroboration, the C layer. If Chosen is low, work on Human Handoff, the H layer: AI referral tracking, human-led sales copy and conversion rate optimization.
What does a typical read look like?
Here is a made-up example to show the method, not a real result.
Picture a business with a strong Found stage and a weak Recommended stage. The site is clean and AI can read it. But when buyers ask the real questions, engines name other businesses. The tier says Ghosted or Friend Zone. The stages say why: the problem is not visibility, it is selection. So the first fix is not a redesign. It is rewriting pages to answer buyer questions and adding FAQ content.
Now picture the reverse. Recommended is decent but Trusted is low. AI likes the answers but cannot confirm the claims elsewhere. The first fix is outside the website: directories, press, reviews and consistent facts.
What should you avoid doing with your score?
Avoid chasing the total when the stage breakdown tells you what to do.
Do not spread effort across all five stages at once. Do not treat a single run as permanent, since engines change. And do not ignore the cap: a business AI never names stays at 49 or below until that changes.
How do you know a fix worked?
Re-score after each major change and see which stage moved.
Because the score uses blind buyer questions across four engines, you can tell a real gain from wishful thinking. The point of the number is to tell you what to do next. It is not a trophy.
Frequently asked questions
What is a good Match Protocol Score™?
Seventy or above puts you in Second Date, which means AI recommends you regularly. Eighty-five and above is The One. The 2026 benchmark average was 47.3, so anything above 50 beats the typical small business tested.
Why is Chosen only 10% of the score?
Chosen measures whether your site converts the human once AI sends them. The other four stages decide whether AI sends anyone at all, so they carry 90% of the weight.
Is the Match Protocol Score™ the same as an SEO ranking?
No. An SEO ranking tells you where a page sits in a list of links. The score tells you whether four AI engines name and recommend your business when buyers ask questions without mentioning you.
Next step
Know your number, then fix the right stage. Book Your Match Protocol™ Strategy Session or Get Your Match Protocol Score™.
Clint Rogers is the founder of Modern Asset Lab and the developer of Match Protocol™.
