Implementation
Work that follows the evidence, not a checklist.
Every intervention starts from a measured gap and ends at a remeasurement. Nothing is done because it is conventional.
The observable foundations
Nobody outside a provider can see how it chooses what to say. What can be observed is the environment the answer is drawn from — whether a team's information exists, is reachable, is consistent, and is supported. These eight areas are where that environment can be improved legitimately, to make a team easier to discover, verify and cite. None of them is a lever on the model.
- Retrieval eligibility
- Pages that can actually be fetched, read and indexed: crawlable HTML, no blocking directives, text that exists in the markup rather than only inside images or scripts. A page that cannot be retrieved cannot be used by anything.
- Topical relevance
- Content that addresses the buyer and seller questions people genuinely ask in that market, in their words, rather than generic service copy.
- Entity clarity
- One official team name, consistently stated brokerage, market, service area and members, so the team resolves to a single identifiable entity instead of several partial ones.
- Factual evidence
- Claims that are specific, checkable and attributed — production, tenure, specialisms, areas served — rather than superlatives no source supports.
- External authority
- Accurate profiles, reviews and legitimate third-party references on sources that already appear in the market's answers. Earned, never fabricated.
- Consistency and freshness
- The same facts stated the same way across every public surface, and kept current as the team changes.
- Extractability
- Information laid out so a reader or a machine can lift a direct answer from it: plain headings, direct statements, structured data where it genuinely matches what the page shows.
- Measurement
- A baseline before the work, a documented change log during it, and a remeasurement on the same locked question set after it.
The work
- 01
Audit the evidence footprint
Establish what exists publicly about the team today, and where.
- 02
Resolve entity inconsistencies
One official name, correct brokerage, correct markets, correct members.
- 03
Improve factual source pages
Accurate, substantiated profile and site content in crawlable HTML.
- 04
Strengthen third-party coverage
Earned coverage on legitimate local and industry sources.
- 05
Improve question-answer content
Direct answers to the buyer and seller questions measured in the market.
- 06
Publish original market research
Verifiable analysis the team can actually stand behind.
- 07
Track every change
Date, owner, rationale and dependency logged for each intervention.
- 08
Remeasure
The same locked question set rerun, results reported against the change log.
This does not replace traditional SEO.
Technical SEO and crawlability remain the foundation. If pages cannot be fetched, read and understood, no amount of evidence work matters. What changes is the measurement problem: classic SEO asks whether a page ranks; this asks whether the team is recommended, cited and described accurately when someone asks an AI who to hire.
The two disciplines share infrastructure and diverge in what they count. We treat them as complements and say so plainly rather than declaring SEO dead.
What this cannot prove
- We measure and improve the evidence environment. We do not control what any AI system outputs.
- No implementation guarantees rankings, recommendations, citations, traffic or revenue.
- Where several changes ship in the same window, single-cause attribution is a hypothesis.
- Results are reported as observed movement against a documented change log.
Keep reading
See who AI recommends in your market — and where you stand.
One client per market. Start by telling me where you work.