Is your website ready for the AI era?

Start with the business question, not the label

An AI-ready website should not be treated as a new product category that every business must buy. A more useful question is whether your website explains your business clearly, can be accessed by the systems you want to reach, and helps a prospective customer take the next step. Those are practical things you can review without predicting which platform will recommend you.

Google says established SEO best practices remain relevant to AI Overviews and AI Mode, with no additional requirements or special optimisations needed for those features. That is a reason to examine the foundations before commissioning an expensive AI-specific rebuild. It is not a promise that a technically sound website will appear in an answer. ^[1]

For a first review, choose three important journeys: finding your main service, checking whether you are suitable, and making an enquiry. Ask someone unfamiliar with the business to complete them. Record uncertainty, missing information and broken steps rather than simply asking whether they like the design.

Separate access from selection

There are several different questions hiding inside the phrase AI visibility. Can a crawler request a page? Can the relevant search system index it? Is it eligible to be used in a particular feature? Does the system actually select it for a particular answer? For Google AI features, a supporting page must be indexed and eligible to appear with a snippet; eligibility does not guarantee inclusion. ^[1]

Ask your developer to review robots.txt rules, indexing directives and any security settings that might block intended crawlers. Review actual responses rather than assuming that a public URL is accessible to every bot. Google specifically recommends allowing crawling and checking that hosting or content delivery infrastructure is not blocking access. ^[1]

OpenAI distinguishes OAI-SearchBot, used for search, from GPTBot, associated with model training. Their permissions are independent. You can allow the search crawler while disallowing the training crawler. Decide what your business wants to permit, document that decision and test the configuration against the current provider guidance. Allowing access is not an endorsement, partnership or recommendation guarantee. ^[2]

Make your commercial information unambiguous

Review your main service pages as if you were a buyer with no prior knowledge. Can you tell what is included, who the service is for, which locations are covered and what happens after an enquiry? Replace broad claims such as complete solutions with a useful account of the actual work. Where scope varies, explain what determines it rather than inventing a universal price.

Google’s people-first content guidance asks whether a page provides substantial information, demonstrates relevant experience and leaves readers with a satisfying answer. Use those questions as an editorial review framework. They do not establish a formula for AI citations, but they provide a defensible basis for improving the page itself. ^[3]

• Explain the service and its boundaries in ordinary language.
• Identify the business responsible for the information.
• Show genuine evidence where permission and confidentiality allow.
• Answer important buying questions, including limitations.
• Give the reader a clear, appropriate next step.

Ask the person delivering the service to review the draft. Their job is to catch overstatements, missing conditions and misleading simplifications. Do not manufacture case studies or add impressive-sounding statistics merely to make a page appear authoritative.

Check the content machines and people receive

Google recommends making important content available in textual form, using internal links to help discovery, and ensuring structured data matches the text visible on the page. Start with those explicit recommendations rather than a speculative checklist of secret AI signals. A beautifully designed page still needs a clear explanation of what it offers. ^[1]

For a practical review, compare the main navigation, service pages and contact details across mobile and desktop layouts. Look for important explanations that exist only in an image or an inaccessible interaction. Ask your developer to demonstrate what the relevant crawler receives where the implementation is complex, rather than relying on a screenshot of the finished page.

Structured data can provide explicit clues about a page’s meaning and support eligibility for appropriate rich results. It should describe real content, not create a second, embellished version of the business for machines. Google states that no special structured data is required for its AI features. Treat markup as a supporting technical task, not an AI visibility switch. ^[1][4]

Keep the website useful after discovery

A review should not stop when someone reaches the website. Test the contact form, telephone links and any booking process. Try the journey with a narrow screen and ordinary business information, including a correction to an earlier answer. Note whether confirmation messages explain what happens next. These are acceptance tests you can specify even without making a claim about conversion uplift.

Performance deserves its own check. Core Web Vitals measure loading, responsiveness and visual stability through Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift. They are useful ways to describe parts of the visitor experience, not a complete definition of website quality. Pair measurement with direct testing of the journeys that matter to your business. ^[5]

Prioritise problems by their consequence. A failed enquiry form should not wait behind a cosmetic animation change. Equally, an attractive new design should not be approved while its service information remains vague. Give each issue an owner, a testable outcome and an agreed place in the work plan.

Turn the review into a manageable action plan

Avoid a single pass-or-fail AI readiness score. Instead, keep separate findings for access, content accuracy, technical implementation and customer journeys. That makes trade-offs visible and helps prevent a broad marketing label from obscuring the actual work. Record what was tested, when it was tested and which evidence supports the conclusion.

Use the following checklist to commission a focused website review. The objective is a clearer, more dependable website with deliberate crawler permissions, not a guaranteed position in an AI answer. Revisit the findings after significant content, platform or security changes, and keep someone responsible for maintaining the information.

• Confirm which search and training crawlers you intend to permit.
• Check important pages for indexing and access problems.
• Rewrite unclear service explanations with the delivery team.
• Validate relevant markup against visible, accurate content.
• Test mobile journeys, performance and enquiry completion.
• Assign owners and acceptance criteria to the priority fixes.