AI SEO: get recommended by ChatGPT, Gemini and Perplexity

How AI answers pick brands

When someone asks an assistant “what’s a good linen blazer for work that doesn’t crease?” or “who’s a good employment lawyer in Leeds?”, the answer usually comes from two places: what the model learned in training, and what it finds when it searches the web in the moment. For buying questions, that live search matters most.

From the sources it finds, the assistant builds a short list. Brands tend to make that list when:

• Several independent sources agree about them. Reviews, editorial articles, forums and comparison pages all saying similar things.
• Their own site states facts clearly. Who the product is for, what it costs, how it compares, what’s included.
• Their product data is complete and consistent. The same price, availability and details in the feed, on the page and in structured data.
• They rank in normal search for the questions the assistant looks up.

Start with what AI already says about you

Before changing anything, we measure. We build a set of 100 to 300 prompts that real customers might ask, across three types:

We run those prompts across the major assistants and record whether you’re mentioned, where you appear in the answer, how you’re described and which sources were cited. That gives us a share-of-answer baseline and, more usefully, a list of the sources the assistants trust in your category.

Make your site easy to quote

Assistants prefer pages that answer a question directly and back it up with specifics. We rework key pages so they:

• Open with a plain answer, then give the detail.
• Use specific facts: materials, dimensions, prices, timings, credentials.
• Include honest comparison tables, including where a competitor might suit someone better.
• Show who wrote or reviewed the content and why they’re qualified.
• Carry accurate structured data for products, organisation, reviews and FAQs.

We also check that your site isn’t accidentally blocking the crawlers assistants use for search, and that key information isn’t hidden behind scripts, tabs or images that crawlers can’t read.

Build the brand footprint off your site

Your own site is only part of the picture. Assistants weigh what others say about you, so we work on the sources they cite most in your category:

Product data for AI shopping

AI shopping features in ChatGPT, Google and others increasingly pull from structured product data rather than reading your pages. For stores, that means your Merchant Center feed, your Shopify catalog and your product schema now affect AI visibility directly.

This overlaps heavily with our agentic commerce work, where we get your catalog ready for AI agents that can shop on a customer’s behalf.

Measuring AI visibility

We report on AI visibility monthly, using measures you can check yourself:

• Share of answer: how often you’re mentioned across the tracked prompts, and how that compares with competitors.
• Citation share: how often your own pages are linked as a source.
• AI referral traffic: visits and sales from assistants, using a custom channel group in GA4 so they don’t vanish into “referral”.
• Branded search lift: people often hear about you from an assistant and then search your name.
• What customers tell you: an “AI assistant” option on your post-purchase or intake survey.

What we won’t do

There’s a lot of nonsense being sold under the AI SEO label. So, for the record, we don’t hide instructions for AI systems in your pages, we don’t create fake reviews or forum accounts, and we don’t mass-produce AI-written pages. These tricks either don’t work or work briefly before they damage your brand.

Questions about AI SEO