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.

Case studiesReal estate

Residential developer, 3 Indian cities · AirPulse analysis

Where a multi-city developer wins AI answers: local budget questions

A residential developer building in Bengaluru, Chennai and Pune gets named in 17.4% of AI answers to budget questions for the localities it builds in, against 0.5% on city-wide questions. On Perplexity its mentions nearly tripled in four weeks. The play is to own the locality and budget questions where buyers are closest to a decision.

17.4%
Named on budget questions in its localities
4.1% to 11.2%
Perplexity answers naming it, in four weeks
0.5%
Named on city-wide questions: the open gap

The challenge

City-wide 'best developer' answers are held by four large names in 98.9% of answers, and the developer's own site supplied 0.1% of citations. Mentions on ChatGPT and Gemini fell between August and September while Perplexity rose.

What we found

Budget questions for localities where it builds named the developer in 46 of 264 answers (17.4%), against 2 of 396 (0.5%) on city-wide questions, Aug 17 to Sep 28, 2026.
Perplexity answers naming it rose from 4.1% to 11.2% in four weeks on the same seven questions.
City-wide 'best developer' answers name one of the same four large developers 98.9% of the time: the set a mid-size developer has to break into.

The play

  1. 1Build locality and budget pages (for example [BHK] in [locality] under [budget]) for every project, where the developer already wins 17.4% of answers.
  2. 2Earn places on third-party 'best developers in [city]' lists and portal developer pages to enter the city-wide set.
  3. 3Track each engine every two weeks, since ChatGPT and Perplexity moved in opposite directions within a month.
  4. 4Keep project facts identical across the site, the RERA record and portal listings.

Results

4.1% to 11.2%
Perplexity answers naming the developer
98 answers per window; 2026-08-17 to 2026-08-30 vs 2026-09-15 to 2026-09-28; same 7 prompts; India

Where AI finds its answers in this category

Listing portals32.2%
Competitors' sites26.5%
Other21.2%
Blogs and guides8.2%
News and media5.2%
Review and marketplace sites2.9%

See how we work with brands like this one: AI visibility for real estate.

How we measured

A residential developer with projects in several Indian cities was evaluating AirPulse. AirPulse tracked seven buyer prompts that do not contain any developer's name, covering city-wide questions and two budget queries for specific localities, once a day on four AI engines in India. The prompt set did not change between the windows compared.

7 tracked buyer questions, 44 daily runs, Aug 17, 2026 to Sep 28, 2026, on ChatGPT, Gemini, Perplexity, Google AI Overviews. Published without the brand's name.

  • Across all seven questions on ChatGPT, Gemini and Perplexity, mentions moved from 12.6% (Aug 17 to 30) to 6.1% (Sep 15 to 28): ChatGPT fell from 11.2% to 1.0% and Gemini from 22.4% to 6.1% while Perplexity rose.
  • The first three days (2026-08-14 to 2026-08-16) were measured without an India region and one prompt was added on 2026-08-16, so all comparisons start on 2026-08-17.
  • A Google AI Overview scraper outage from 2026-09-12 to 2026-09-28 affects the later window, so mention rates exclude Google AI.
  • Naming is measured by text match on the answer text. Rates rest on 98 answers per engine per window, so a few answers move a percentage noticeably.
  • Source categories are assigned by domain. Broker and channel-partner project sites count as listing portals; 'Other' is mostly Google Maps and search links and image hosts.
  • This was an evaluation. No work by AirPulse on the developer's site is documented, and the data does not show why mentions fell.

See where AI recommends you.

Book a free strategy call and we run your buyers' questions on the engines while you watch.