Start with the answer, not the dashboard
I ask the buyer’s question in the same systems the buyer uses. I save the language, the citations, and the companies named. That record is more useful than arguing about a visibility score.
Why I wrote it
I kept sending founders the same screenshot.
They knew their business. They had a decent site. Some ranked well. Then they asked ChatGPT, Gemini, Perplexity, or Google the question a buyer would ask—and watched a competitor get the recommendation. The Screenshot is the field guide I wanted to hand them next.
THE DISTINCTION
Ranking and being recommended are related. They are not the same job.
Search engine optimization (SEO) is still the foundation: a site has to be crawlable, useful, technically coherent, and relevant to the query. Generative engine optimization (GEO) begins where the blue-link checklist ends. It asks whether an answer engine can resolve the company, trust the description, find a source worth citing, and state the difference in one clean sentence.
I do both through Suede SEO because the evidence crosses the boundary. A canonical problem can weaken rankings. A thin entity record can erase a company from an AI answer. A strong original page can serve the reader, the search result, and the model that later quotes it.
WORKING NOTES
The method in three moves
Follow every description to its source
If an answer gets the company wrong, something upstream usually made the wrong version easier to believe: a vague page, a split entity record, stale third-party copy, or a stronger competitor source.
Fix the record in public
The durable repair is material another person can open and check: a clear page, a dated report, a real author, a canonical URL, useful structured data, and independent coverage where it belongs.
“The screenshot does one useful thing: it ends the argument about what the machine said.”
From the author’s working premise
The complete work
Eleven chapters, a glossary, and no gate on the web edition.
Read it in the indexed edition at Suede SEO. The public GitHub edition documents the build and provides a clean reading copy. Version 1.0.0 includes the PDF and EPUB.