01What happens.
A website can serve people in full and turn AI away. The public site loads, conventional search indexes it, and the digital measures a leadership team relies on read as normal. The routes AI uses to reach the same site receive something different: some are served, some are challenged, some are turned away and some receive no reply, and none of that shows in an ordinary check.
02Why it matters.
If you are the chief financial officer signing off spend on AI visibility, the question is whether the systems the money is meant to influence can reach the material being promoted. Where they cannot, the spend goes to third-party material about the organisation, because that is what the systems can read. 85% of brand mentions in AI search come from third-party content, not the brand's own words, and restricting access to the organisation's own content raises that share.
The door changes without notice: a security setting, a network change or a new crawler policy can close it silently. The visibility programme will not detect the change, because it measures what AI says and not what AI can reach.
03The case.
The organisation is anonymised, the finding is AAAnow's own, and the evidence base is set out in AAAnow's published doorway assessments.
The organisation had a problem it could see: AI answers about it were thin, out of date and drawn from other people's material. It bought a programme to fix that. The £350,000 covered content, structured data, promotion and monitoring over a year, measured by what AI said about the organisation each month.
AAAnow assessed the organisation's digital doorway from the outside, the position a customer, a search engine or an AI system stands in. The 4 tested routes used by people were served. Of the 15 tested AI-related routes, most were turned away or received no reply. The site was open to people and, for most of the tested routes, closed to the systems the programme was paying to reach.
Nothing in the programme's monthly reporting showed it. AI answers about the organisation were changing slowly, and the changes were being attributed to the programme. The answers were assembled from third-party material, older pages and whatever was reachable, because the organisation's own current material was not, and the programme was promoting a position the systems could not verify at source.
Across 5 anonymised doorway assessments AAAnow published in Aug. 2026, people were served in full in 4 of the 5 while between 4 and 14 of the 15 AI-related routes were turned away or received no reply. Each organisation appeared fully operational through the checks its leadership relied on.
The chief financial officer's question, once the assessment was on the table, was what the £350,000 had bought. The answer was a better version of a message that most of the tested AI routes could not receive, and a year in which the foundations the message rested on had not been looked at.
04Exposure.
The spend is the visible exposure: a programme whose value depends on access the organisation had unknowingly withdrawn, and whose reporting could not show the withdrawal. The wider exposure is what filled the gap. Where the organisation's own current material was unreachable, historic pages, third-party commentary and archived documents acquired greater influence over the answer, and the organisation's brand investment was working against material it did not control.
The board's interest is in whether investment in content, digital infrastructure and AI visibility can produce value when the authoritative information is inaccessible, and in whether each restriction at the doorway is known, intentional and owned. In this case none of the restrictions had been decided by anyone. They had accumulated, and the visibility budget had been spent on top of them.
05What AI Readiness changes.
What the organisation could not see from the inside was that its doorway was turning away most of the AI routes it was paying to reach, because from the inside the site worked. AAAnow tests the routes AI uses from the outside, 15 of them, and evidences whether each is served, challenged or turned away, on the day of assessment and continuously after it. With that evidence the board can decide which restrictions it wants, record the reason for each, and put the foundations in place before the next pound is spent on being seen.
Your team makes the changes; AAAnow provides the assessment, the evidence, the guidance and the review.
06Related use cases.
A hostile version of the results was the one AI quoted, because the company's own was locked away
The same closed doorway on results day.
Announcements buried in the page furniture
Visibility spend that amplified a problem nobody had diagnosed.
2 AI tools, 2 different answers, and consequences the organisation could not see in full
The doorway as one contributor to different answers from different tools.
07References.
- £350,000: AAAnow observed, anonymised.
- 5 anonymised outside-in doorway assessments, Jul. and Aug. 2026: Lawrence Shaw, Digital Suicide, LinkedIn, 3 Aug. 2026.
- 85% of brand mentions in AI search come from third-party content, not the brand's own words: AirOps, 21,000+ brands. www.airops.com/blog/llm-brand-citation-trackingairops.comWe do not enable external links - please copy and paste, you are then certain as to the link being opened
Next step
Talk it through, in your terms.
The briefing sets this case in your organisation’s terms, and the discussion follows your questions.
Observed, anonymised. No company is identified.