AI Kick Start — Free Template
Generative Engine Optimisation (GEO) is how you get ChatGPT, Gemini, and Claude to understand your business and recommend it when a customer asks. Work through this checklist to find the gaps stopping AI engines from citing you.
How to use this: tick a box once your site genuinely meets the item — not when you intend to. Print it, walk it page by page, and note the URL where each gap lives. Anything left unticked is a job for your next content sprint.
AI engines build a mental model of "who you are". Mixed names, addresses, and descriptions across the web make that model fuzzy and unreliable.
Your legal and trading name appears the same way on your site, Google Business Profile, social profiles, and directories — no abbreviations or stray suffixes.
Name, address, and phone are consistent everywhere, including state and postcode format for your Australian location.
You have a single, plain-English "what we do and who for" sentence, used verbatim in your homepage, About page, and profile bios.
The suburbs, cities, or regions you serve are named in words on the page, not just implied by a map embed.
Schema gives engines machine-readable facts they can quote with confidence. Missing or broken schema means they have to guess.
Your homepage carries valid JSON-LD with name, URL, phone, address, and areaServed. It validates with no errors in a schema checker.
Key offering pages use Service, Product, or Offer schema so engines can match them to specific customer questions.
Pages with genuine question-and-answer blocks expose them as FAQPage schema, so the answers can be lifted directly.
Every fact in your structured data also appears in the page text. No hidden or contradicted claims that could be flagged as deceptive.
Engines quote content that is easy to extract: clear definitions, hard numbers, side-by-side comparisons, and direct answers. Walls of marketing copy get skipped.
Important concepts open with a "X is …" sentence an engine can quote without rewriting.
You include real numbers — prices, timeframes, volumes, outcomes — stated as facts an engine can cite, with a source or date where relevant.
Where customers weigh options, a clean comparison table lays out the factors. Tables are highly quotable and map well to "X vs Y" questions.
You answer the questions customers actually ask, in their words, with the answer in the first sentence and detail after.
Section headings read like the queries people type, so engines can match a heading to an intent at a glance.
You cannot improve what you do not measure. Knowing whether engines currently mention you is the baseline for every other item here.
You have a list of the questions a customer would ask an AI engine to find a business like yours.
You have run those prompts in ChatGPT, Gemini, and Claude and recorded whether you are mentioned, cited, or absent.
Your robots.txt does not block the AI engine crawlers you want citing you, and your key pages return clean, crawlable HTML.
You have a recurring date to re-run the prompts and track whether visibility is improving after changes.
Engines cross-check your claims against the wider web. Independent mentions, reviews, and citations build the trust that gets you recommended.
You appear in relevant Australian directories and industry listings with consistent details.
You have genuine reviews on platforms engines read, with enough volume and recency to look active.
Articles, partner pages, or press describe you the same way your own site does, strengthening the entity model.
Real people, credentials, and experience are shown on the site, supporting the E-E-A-T signals engines weigh.
Mostly ticked: Your fundamentals are sound. Focus on tracking and off-site signals to widen your lead.
Half ticked: The usual gaps are schema and citable formats. Fix those first — they pay back fastest.
Mostly empty: Start with entity consistency and a baseline visibility check before producing more content.
Tip: use your browser's Print command (Ctrl+P / Cmd+P) to save this as a clean A4 PDF or hard copy.