Get Recommended by AI, the book by Ronnie Huss

Get Recommended by AI

A practical book about how AI search engines decide which businesses to put forward, and what you can actually change about it. It is written for founders, marketers and consultants who have noticed that ChatGPT, Claude, Gemini and Perplexity now answer the questions their customers used to type into Google, and who want to know whether they are in those answers.

It is 219 pages, 36 chapters across 8 parts, and it is built on the same measurement that produces the SearchScore benchmarks: a corpus of more than a million audited websites.

What it covers

  • Part 1, The Shift. How an AI answer is actually assembled, the difference between what a model remembers and what it retrieves, and what a million websites reveal about who gets named.
  • Part 2, Diagnose. Running your own visibility baseline, the competitor replacement test, and how to read the result honestly.
  • Part 3, Foundation. Crawler access engine by engine, llms.txt as cheap insurance rather than a magic bullet, structured data without the myths, entity clarity, and a 15 minute Wikidata walkthrough.
  • Part 4, Extraction. Answer-first writing, the five blocks, query fan-out, and why one page is never enough.
  • Part 5, Reinforcement. Why third parties decide, how to find yours, becoming the stat source, author authority, and where AI actually gets its sources.
  • Part 6, Track and repair. Volatility and drift, building a tracking system, objective measurement using server logs and AI referrals, and the repair playbook for when AI gets you wrong.
  • Part 7, Execute. A 7 day starter sprint, a 30 day plan, the priority matrix, the eight common mistakes, and the shift to AI that acts rather than only answers.
  • Part 8, The toolkit. The templates, schema blocks, worksheets and query bank, ready to use.

Why I wrote it

I kept running the same audit and finding the same thing. Businesses were not being left out of AI answers because they had blocked the crawlers or because their websites were slow. They were being left out because nothing on the page said, in a form a machine could read, who they were and what they knew.

That is not an opinion. Across 9,487 UK law firm websites, three were ready to be recommended. Across 1,038 accountancy firms, not one wrote answers in the format AI quotes from. In the most recent reading of the full corpus, 216 sites out of more than a million reached the AI-Ready tier and the highest score recorded anywhere was 86.6 out of 100.

Articles were not the right format for that. A method was.

The evidence is published in full

Every figure in the book comes from research that is public, with its method and its limits attached. You can read the index, the methodology and the measurement standards without buying anything, and my own reading of the findings is in the research.

What comes with it

Every reader gets the digital toolkit, which is the templates, schema blocks, worksheets and Notion databases from Part 8, plus twelve months of SearchScore Tracker on the Founders tier. The access code is printed inside the book, and you claim both at searchscore.io/unlock.

Where to get it

It is out now in Kindle, paperback and hardcover. The book has its own site at getrecommendedbyai.co.uk, where every chapter is listed, along with a glossary of the vocabulary, a free chapter and the press kit. Get your copy here. That page shows the current price and sends you to your own Amazon store, so it works wherever you are reading this from.

If you would rather not buy a book

Everything in it is downstream of the same idea, and the free version starts here: the guide to generative engine optimisation, and the research.

About the author

Ronnie Huss, founder of SearchScoreRonnie Huss is the founder of SearchScore, an AI visibility platform that has audited more than a million websites. He created SAVI, the State of AI Visibility Index, and has published sector readings for law firms, accountancy practices, dentists, veterinary practices, care homes and aesthetic clinics. He publishes SearchScore’s methodology and measurement standards in full, including where its own figures are directional rather than decision grade. He is a serial founder with product launches across SaaS, AI tools and blockchain, is based in London, and works with clients internationally.

For media and events

Review copies, interviews and speaking: get in touch. The SearchScore press kit has boilerplate, figures and logos.