Digital Elevator · AI visibility baseline · 2 September 2026
The opening measurement for the Managed SEO programme. 200 probes — 40 buyer prompts across five AI surfaces — scored for whether an engine names Global App Testing, where in the answer, and how it speaks about the company.
Share of all 200 probes in which each brand was named. Fixed for the whole programme, so the next read is directly comparable.
Global App Testing’s visibility on each engine, across all 40 prompts.
All 40 tracked prompts and what each engine did with them. Filter to compare how Global App Testing performs by surface and by the kind of question being asked.
| Prompt | Type | Engines | Hit | Rivals named |
|---|
Engines assemble answers from sources. 100 distinct domains appeared across this run — which ones, and who owns them, decides who gets named.
What a probe is. One prompt asked of one engine. 40 prompts across five surfaces gives 200 probes. Visibility is the share of those in which the company was named at all.
The prompt set is fixed. It stays the same every period, because a percentage only means something against a stable denominator. It was built from the queries the AI search visibility audit tested, weighted toward questions where engines actually recommend suppliers rather than explain a concept.
Share of a tracked set, not of the market. Rank and share here are against the six competitors chosen for this programme. Adding a seventh would lower the percentage without anything changing in reality, which is why the raw count is shown alongside.
Five surfaces, not six. ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews / AI Mode. Microsoft Copilot cannot currently be measured — no data provider exposes it — so it is excluded rather than estimated. Additional surfaces can be added and are quoted separately.
This is not a ranking guarantee. It is a measurement of the before-state, and the input that decides which pages, which comparisons and which technical specifications get built first.