search is becoming agentic

Agentic Search Optimisation: get your brand chosen by AI agents

AI agents in ChatGPT, Gemini and Perplexity no longer just answer questions. They research, compare prices and availability, check reviews and, more and more often, complete the booking or checkout for the user. I help make sure your website is one they can find, understand, trust and act on.

What is agentic search optimisation (ASO)?

Agentic search optimisation is the practice of structuring your website content, data and trust signals so that autonomous AI agents can discover, evaluate and select your business on a user's behalf. You'll also hear it called agentic SEO or AI agent optimisation. (It's not the same as app store optimisation, which also uses the ASO acronym.)

SEO vs GEO vs agentic search optimisation

SEO GEO / AEO Agentic search optimisation
Who's searching A person using Google or Bing A person asking an AI assistant An AI agent acting for a person
Goal Rank and earn the click Get mentioned and cited in the answer Get shortlisted, chosen and transacted with
What wins Relevance, links, technical health Clear, citable content, consistent entity information, third-party mentions Machine-readable pricing, availability and policies, verifiable proof, and pages, forms and checkouts an agent can use
How it's measured Rankings, organic traffic, conversions Brand mentions, citations, AI referral traffic Agent shortlisting, task completion, assisted conversions

ASO builds on SEO and GEO rather than replacing them. If your pages can't be crawled and AI platforms never cite your brand, an agent won't find you in the first place. That's why every ASO project starts with technical SEO and AI Overview and LLM optimisation.

How AI agents decide which businesses to recommend

A useful way to think about it is the three-stage model popularised by Promptwatch and First Page Sage. Tools like ChatGPT agent and agentic browsers such as Perplexity Comet tend to move through these stages:

  1. Retrieval: The agent splits the user's goal into several searches. It builds a shortlist from what it can crawl and from what it already 'knows' about each brand.
  2. Evaluation: It compares the shortlist on hard facts: pricing, availability, delivery and returns, reviews and independent mentions. Brands whose information is missing or contradictory are more likely to be dropped.
  3. Action: It fills in the form, books the slot or completes the checkout. If a machine can't operate your booking flow or checkout, the task can fail, or the agent can move on to a competitor whose flow works.

From invisible to agent-ready in three steps

Agent visibility & readiness assessment

I benchmark how ChatGPT, Gemini, Perplexity and Google AI Mode currently find and describe your brand compared with your competitors. I also audit how ready your site is, technically and commercially, for AI agents.

Prioritised action plan

You get a bespoke report with every issue ranked by impact and effort. It covers everything from quick structured data and crawler fixes to longer-term proof-building and protocol work.
P

Implementation & tracking

Take the plan to your in-house team, or hire me to implement it with them. Either way, we'll re-test agent journeys and track AI visibility and conversions over time.

‍

get up to speed

Improve decisions with data

Agent readiness audit

First, the basics agents depend on: whether AI crawlers such as OAI-SearchBot, GPTBot and ChatGPT-User can reach your pages through your robots.txt, CDN and firewall rules, whether your content renders without JavaScript.

Machine-readable commercial data

Agents shortlist on facts. I make sure your pricing, availability, product and service details, delivery and returns are published in clean, semantic, accessible HTML and Schema.org structured data, and that they match across your website, product feeds and Google Merchant Center.

Trust, proof and entity consistency

Agents check your claims against reviews, third-party mentions and other sources. I audit how AI platforms describe your brand (see brand accuracy audit), find contradictions and gaps, and build the proof that gets you through the evaluation stage.

Agent journey testing

I run your key journeys through AI agents, from 'find me a…' through to completing a form or checkout, and record where they get stuck. That includes whether buttons and form fields are labelled clearly enough for a machine to use.

LLM Search Optimisation FAQs

Is agentic search optimisation different from GEO?

Yes, although the two overlap. GEO focuses on getting your brand mentioned and cited in AI-generated answers. ASO also covers what happens next: whether an AI agent can compare you on price and availability, verify your claims and complete a task such as a booking or purchase on your site. Most businesses need both, and good SEO underpins each.

Which search engines do AI agents use?

It depends on the assistant. Gemini and Google AI Mode use Google, and Microsoft Copilot uses Bing. ChatGPT runs its own crawler (OAI-SearchBot) alongside third-party search providers. Claude reportedly uses Brave Search, and many custom-built agents call search APIs such as Exa or Tavily. So being indexed and accessible beyond Google, starting with Bing, is part of agentic search optimisation.

Do you analyse server logs for AI agent traffic?

Yes, wherever you can access them. Logs are the most reliable record of AI crawlers and agents visiting your site, because many never trigger analytics tags. I work with raw server logs or CDN logs from providers such as Cloudflare, Akamai and Fastly. Some hosted platforms, such as Webflow, Shopify and Wix, offer little or no log access. In that case I use CDN data where it's available and rely more on prompt tracking and agent testing.

What's included in an agentic search optimisation service?

Usually an AI visibility benchmark across ChatGPT, Gemini, Perplexity and Google AI Mode, a technical agent-readiness audit, a review of your structured data, product feeds and commercial information, brand accuracy checks, agent journey testing and a prioritised action plan. Implementation and ongoing tracking are optional.

How long does it take to see results?

Technical fixes such as crawler access and structured data can start to take effect once AI platforms re-crawl your pages. Changing how AI platforms describe and recommend your brand usually takes longer, because it depends on third-party mentions and how often models refresh their sources.

How should I choose an agentic search optimisation consultant?

Look for someone who can show you how they test AI agents on real journeys, how they measure visibility and agent access (not just rankings), and who is honest about which emerging standards are worth your time.

get in touch

Let's improve your website!

Google rewards high-quality websites and there isn't a shortcut to success. Shall we start with a comprehensive, actionable site review?

Your message has been submitted.
I'll get back to you within 24 hours.
Oops! Something went wrong. Please try again!