Lesson 2 of 5 · 3 clips · 11 min
What makes AI recommend a brand
When an AI answer recommends three companies, what do those three have that the others do not?
Three clips, two guests, one thread: the answer is assembled from everything said about you, on your site and off it, and a gap anywhere shows.
Clip 1: Work backward from the brands that already show up
Who is speaking: Solomon Thimothy.
Solomon Thimothy's method: run the prompts your buyers would use, note which brands appear, then go and find the links and mentions those brands have. His read is that two things from early SEO still decide it: good links and frequent mentions of the brand in different places. He also sees the engines disagree. Appearing in Google's AI Overview does not mean appearing in ChatGPT or Perplexity for the same prompt.
Where our evidence differs: He acknowledges that answers are personalized, then works from one run of each prompt as the answer everyone gets. When we asked the same questions repeatedly, the same question asked twice did not reliably return the same sources, so read any single run as one sample. SEO 101, lesson 3: how AI engines decide what to cite.
Clip 2: Buyers check you before they call you
Who is speaking: Matt Bertram until about 2:59, Corey Morris until about 4:06, then Matt.
Matt's point is about the research step. However a buyer first hears of you, they go and check, and in Similarweb data he now sees AI tools among the sources sending people to brand sites. Corey Morris draws the content consequence: a page that gets one visit a year may still need to exist, because it covers a service you offer or the one prospect who fits, so keep it updated and accurate rather than pruning it or letting it go stale. AI, he says, judges the complete picture even more than Google did.
Clip 3: Volume will not tell your brand story
Who is speaking: Matt Bertram until about 0:43, Corey Morris until about 3:46, then Matt.
Matt opens on disambiguation: be clear enough about who you are that an engine neither confuses you with someone else nor misses that your profiles are one company. Corey Morris argues that shrinking traffic has pushed marketers back to what they always said mattered: brand strategy, messaging and product development. In his process those are prerequisites, and mass produced AI content cannot stand in for them. Matt closes on diagnosis: before buying SEO or ads, find out whether the real problem is the website, the brand, or a lack of demand.
Try this before the next lesson: Ask ChatGPT and Perplexity what your company does and who it serves. Wherever the answer is wrong or thin, look at the sources it lists. The page or profile behind the worst answer is your first fix, whether it is yours or someone else's.
All 5 lessons · SEO 101, the companion course for business owners