7 AI Search Optimisation Mistakes That Kill Your

1. Only optimising for Google

Google still leads traditional search. But AI answers pull from a wider set of sources. ChatGPT relies on Bing's index. Perplexity runs its own crawler. If you only optimise for Google, you are invisible to AI engines that use different data. Make sure Bing indexes your site. Set up Bing Webmaster Tools. Check you are not blocking non-Google crawlers.

Fix: Verify your site in Bing Webmaster Tools. Check that your robots.txt allows Bingbot, PerplexityBot, and other AI crawlers. See our guide to structuring your site for AI crawlers.

2. Blocking AI crawlers without knowing it

23% of UK sites block at least one AI crawler in their robots.txt, often without knowing it. Some WordPress security plugins and CDN settings block unfamiliar bots by default. If GPTBot, ClaudeBot, or PerplexityBot cannot reach your content, you cannot appear in AI answers.

Fix: Audit your robots.txt and server access logs. Check our AI crawler directory for the full list of bots to allow.

3. No Organisation schema

44% of UK businesses have no Organisation schema on their website. Organisation schema tells AI engines who you are, what you do, and where you are. Without it, AI platforms guess. They often guess wrong. Add Organisation schema with your company name, description, address, and social links. This gives AI engines a reliable source of truth about your business.

Fix: Add Organisation schema (JSON-LD) to your homepage with name, description, url, logo, address, and sameAs links to your social profiles.

4. Vague H1 headings

53% of UK SME websites use H1 headings that do not state what the page covers. Headings like "Welcome" or "Home" tell AI nothing. AI platforms use headings as key signals for page content. A clear H1 with your primary topic makes it much more likely you will be cited.

Fix: Rewrite every H1 to be a clear, specific statement of what the page covers. "Welcome to Our Agency" becomes "AI Search Optimisation Agency for UK B2B Companies."

5. Allowing training bots to scrape your content for free

There is a difference between letting AI search engines cite your content and letting training bots scrape your site to build their models. You want the first. You may not want the second. Bots like GPTBot (for training) and CCBot harvest content to train models. If you give away your content for training with no visibility in return, you help your competitors for free.

Fix: Decide which bots to allow and which to block based on whether they provide search visibility in return. Our AI crawler directory explains the purpose of each bot.

6. Not updating content monthly

76% of AI-cited pages were updated within the last 30 days. AI engines favour fresh content. If your key pages have not changed in six months, fresher pages are beating them. You do not need to rewrite everything each month. Review data, update stats, add new examples, and refresh dates where the content has genuinely improved.

Fix: Set a monthly review schedule for your most important pages. Update statistics, add recent examples, and ensure your content reflects current information.

7. Ignoring YouTube

YouTube presence has a 0.737 correlation with AI search citations. That makes it the strongest single factor. AI engines cite brands with YouTube presence more often. YouTube gives AI engines transcripts, visual context, and an extra source to check your expertise. No YouTube channel means you miss the strongest signal.

Fix: Create a YouTube channel and publish content that mirrors your key website topics. Even basic explainer videos with clear titles and descriptions improve your AI visibility profile. Read more in our analysis of why YouTube is the strongest AI search signal.

What to do next

Start with your robots.txt and schema markup. These are the quickest fixes with the biggest impact. Then work through the rest in order of effort. For a full approach, see our guide on whether AI SEO works and our recommended tools for tracking progress.