Site index for AI agents: /llms.txt. Documentation pages under /docs are also served as markdown at the same URL plus .md (e.g. /docs/geo-audit.md) or via an Accept: text/markdown request.
Dinner · Bengaluru · Sep 2We’re hosting round 3 of How to Destroy Your Competitors on AI Search, a 25-seat dinner.Request an invite →
Case studies

Dated results you can forward.

Stable prompt list. Real fetches. If we changed the measurement, we say so.

EdTech
54% 97%
Google AI mention rate on the tracked prompt set, May–Jun 2026

Newton School barely appeared when buyers asked AI about online tech education. AirPulse listed the prompts, tracked named or not named engine by engine, and re-ran the measurement after each fix until the answers changed.

Named or not named tracked per prompt, engine by engine
llms.txt fixed and verified live
Structured data shipped across key templates
Property inspection
+54%
citations, week over week on the same prompt set, 2026

AirPulse tracked the prompts home buyers actually ask and found the citation gaps. The content and schema fixes that followed grew PropChk's cited answers within days — cited, not yet named, and we report the difference.

Week-over-week citation growth measured on a stable prompt set
Competitor movements monitored so wins do not quietly reverse
Fintech
1 audit
found the indexing gap ad spend was hiding, 2026

BytePe was investing heavily in performance marketing while a broken JSON-LD setup kept pages from being read properly. One AirPulse audit caught it. Their blog now runs on a hosted setup our team built and maintains, matched to their design system.

Schema errors identified and prioritized in the first audit
Blog set up by our team on their domain, no engineering sprint on theirs
Content pipeline connected to tracked buyer prompts
B2B SaaS · Events
#4 #3
webinar citation rank across tracked prompts, three weeks, 2026

AirPulse runs on Airmeet's own marketing every day — the same loop we sell, used on ourselves first. Missed opportunities become briefs, briefs become drafts, and every fix is re-measured on the same prompts.

100+ buyer prompts tracked across 4 engines, daily
Citations and share of answer benchmarked against Zoom, ON24 and more
Weekly digest keeps one marketer on top of the loop
Our evidence standard

Why our numbers look smaller than the ads.

We only publish a week-over-week change if the prompt set stayed still. We say when we cannot prove cause.

Want your brand on this page?

Start with the free check: named or not named on your domain, in 3–5 minutes. Then we walk you through the first four weeks.