Query–passage similarity
cos(q, p) = (q · p) / (‖q‖ ‖p‖)
How closely your best passage matches each buyer query in meaning, against the best competitor passage.
Diagnosis · Retrieval Opportunity Report
A mathematical roadmap for the service that matters most.
We model how Google and AI systems are likely to retrieve one of your most valuable service categories, measure where competitors outscore you on standard retrieval measures, and set out what to change, in order of commercial value.
The retrieval problem
Most visibility advice is opinion. Retrieval can be measured.
Search engines and AI systems decide what to retrieve with scoring functions: how closely a passage matches a query, how clearly an entity is defined, how well a page is connected. Their exact functions are not published, but the standard measures they are built on can be calculated for your pages and your competitors' pages, against the queries your buyers actually use. The report does that for one service category, so every recommendation comes with the number it is meant to move.
What we measure
The numbers this work is built on.
Each is calculated for your pages and your competitors', against the same buyer queries, before anything is changed.
Lexical relevance
BM25(q, d)
The term-weighted match between each query and each page: the scoring family behind keyword retrieval.
Query coverage
|{q ∈ Q : max sim(q, p) ≥ τ}| / |Q|
The share of commercially important queries for which you have a passage above the retrieval threshold.
Entity completeness
attributes stated / attributes expected
How fully your services, locations, people and credentials are stated as explicit, connected facts.
Answer share
k / n
In how many of n sampled AI answers to a query your business is named.
The work
What we do.
Query set
Build the set of commercial queries for the category from search demand, buyer language and the questions your enquiries actually contain.
Passage index
Index your relevant pages and your main competitors' at passage level, so every page is scored on the same terms.
Scoring
Score every query against every passage with lexical and semantic measures, and sample AI answers to record who is named.
Gap analysis
Find where competitors outscore you, and the missing facts, entities or page structures that explain each gap.
Prioritisation
Rank every fix by the commercial value of the queries it affects and the score change it should produce.
Deliverables
What you receive.
Entity and ontology assessment
How your business, services and their relationships are currently defined, and what is missing.
Commercial query and intent analysis
The queries that matter for the category, grouped by intent and weighted by value.
Core service page review
Passage-level scores for your main pages against the query set.
Competitor comparison
The same measures, applied to the competitors you lose work to, side by side.
Retrieval deficiencies
Every point where your information fails to match, connect or be understood, with its cause.
Prioritised opportunities
The changes, ranked by expected score movement and commercial value.
Technical and content recommendations
Specific changes to pages, structure, markup and copy.
Implementation roadmap
The sequence of work, the measure each step should move, and how it will be re-tested.
How it starts
From application to roadmap.
-
01
Apply
Answer six short questions about your business. We read every application.
-
02
15-minute consultation
Free. We show you two or three verified retrieval weaknesses in your market.
-
03
Retrieval Opportunity Report
A measured diagnosis of one service category, and a prioritised roadmap.
-
04
Implementation
The roadmap carried out, then re-tested against the same measures.
Common questions
Is this an SEO audit?
No. An SEO audit checks a site against a list of good practice. This report scores your information against specific buyer queries with the measures retrieval systems are built on, compared with your competitors, and tells you what to change to move those scores.
Do you know how Google or ChatGPT rank?
Nobody outside those companies does: their ranking is not published. We calculate the standard retrieval measures (BM25, embedding similarity, link-graph PageRank) with our own models as a close stand-in, and check them against what the systems actually answer. Answer share is observed, not modelled.
Why one service category?
Because a measured diagnosis has to be specific. One commercially important category, analysed properly, produces recommendations you can act on and results you can check. Further categories can follow.
Do you guarantee rankings or enquiries?
No. Nobody controls how Google or an AI system ranks. We measure the factors retrieval systems are built on, show what each change is expected to move, and re-measure after the work.
What do you need from us?
The pages in scope, the competitors you lose work to, and agreement on which service category matters most.
How do I get one?
Apply for the free 15-minute consultation. If the report is the right next step, we will explain it on the call.
The rest of the roadmap
Free 15-minute consultation
Find out where your business stands.
- How search and AI systems currently interpret your business
- Where competitors are better understood than you
- Specific retrieval weaknesses, checked before we speak
- Whether there is work worth doing, and what it is worth to you
Apply for a free 15-minute consultation
We reply within one business day, and usually within the hour.