AI Search Optimisation: Get Recommended by ChatGPT

Nodus Ai systems · Published 21 August 2026

Your customers have started asking ChatGPT instead of Google. When someone types "best plumber near me" or "which CRM should a small trades business use" into an AI assistant, the answer comes back as a recommendation — a handful of named businesses, not ten blue links. If your business isn't one of the names, you're invisible in a channel that's growing every month. AI search optimisation is the work of becoming one of the names.

What is AI search optimisation?

AI search optimisation (you'll also see answer engine optimisation — AEO — and generative engine optimisation — GEO, all describing the same territory) is the practice of making your business findable and citable by AI systems: ChatGPT, Perplexity, Google's AI Overviews, Copilot. Where traditional SEO earns a ranking, AI search optimisation earns a mention — being the business the engine names when someone asks for a recommendation.

The two are related but not the same. AI engines lean on many of the signals Google uses, but they weight them differently, and they add requirements of their own.

How AI engines decide who to recommend

From what's observable across engines, recommendations draw on a few consistent sources:

How to check where you stand right now

Ask the engines yourself — it's free and takes ten minutes. Put the questions your customers would ask into ChatGPT and Perplexity: "best [your trade] in [your city]", "[problem you solve] — who should I use?". Note who gets named, and whether you do. Then look at what sources the engines cite — those pages are the ones deciding the answer, and being on them is the most direct move available.

What actually moves the needle

In rough order of effort-to-impact for a small business:

  1. Schema markup on your site — Organization, LocalBusiness, FAQPage. This is table stakes and mostly a one-off.
  2. Get onto the cited pages — the directories and lists the engines already draw from. Genuine listings only; the same purchased-link shortcuts that damage SEO damage this too.
  3. Answer-shaped content — pages built around the real questions buyers ask, answered in the first paragraph (the same discipline behind a good AI employee answering a caller directly) rather than after 800 words of preamble.
  4. Consistency sweep — one name, one description, one set of details, everywhere.
  5. Measure on a cycle — re-ask the same questions monthly and track whether your mention rate moves. This is measurable, which means it's improvable.

Does traditional SEO still matter?

Yes — and this is worth saying plainly, because "SEO is dead" makes good headlines and bad strategy. AI engines cite pages that rank; authority still compounds; a site invisible to Google is usually invisible to the engines too. The practical shift is that SEO is no longer the whole game: ranking and being citable are now two halves of the same work, and the businesses doing both are the ones that show up in either channel.

Doing it yourself vs using a service

Everything above can be done by hand. The catch is cadence: AI answers shift as engines update, and a one-off effort decays. The work is checking your mention rate, refreshing answer content, and maintaining listings — monthly, indefinitely. Whether that's a job for you, a marketer, or an automated service comes down to the same test as any other business system: will it actually happen every month if it depends on you remembering?

Related reading