I think marketing leaders who want their brand named in AI answers should spend the first dollar on the coverage, reviews and creator content other people publish about them, and the second on rewriting their own website.
The data that changed my mind
Cision’s State of Search and Social 2027 report, released on November 9, draws on 564.61 million social media posts and 4,678 search queries and AI prompts. One finding stands out for anyone with an AI visibility budget. In AI responses to “best car” queries, Toyota led with a 13.8% share of brand mentions. Toyota’s own website accounted for a fraction of the sources the models cited. YouTube was cited more than 4,000 times.
Cision adds that the most-cited sources in that analysis were YouTube creators, editorial brands and individual journalists, ahead of the carmakers. The release puts it plainly: “The coverage, reviews, and community conversation a brand earns are now the raw material for the answers its buyers receive.” Amy Jones, Chief Marketing Officer at Cision, said: “Search can show where interest is changing. Social helps explain why. And AI answers reveal the narrative reaching consumers.”
A second finding points the same way. For magnesium glycinate, Cision found that stress relief dominated social discussion of the supplement’s benefits, while AI responses emphasized bone support and cardiometabolic support. The answers people get can frame a product differently from how buyers talk about it. A brand that only edits its own pages has little say over either.
The case for starting on your own site
The strongest counter-argument deserves a fair hearing. A brand’s own site is the one source it fully controls. Product specs, prices, availability and policy details live there, and when an AI answer gets a fact wrong, the fix often starts with a clear page that states the fact. Structured content and clean product pages are cheap compared with a creator programme, and they help classic search too. There is also a fair objection to the data itself: Cision sells social listening and media intelligence, and a report that finds earned media drives AI answers supports its own products. And one category of consumer cars is not every market.
I accept all of that, and it still does not change the order of spending. The Toyota numbers show what happens when a brand does well in AI answers: it led its category on mentions while its own site was a minor citation source. Owned pages decide whether the facts are right. Third-party sources decide whether the brand shows up in the answer at all, and how it is framed. If I had to fund one of those first, I would fund the one that controls inclusion.
B2B review platforms are making the same argument from the other side. G2 said on November 10 that AI answer engines base their answers on credible, third-party evidence, and it now sells paid tactics to influence which brands appear in them. G2 has a commercial interest in that view as well. Two companies with different products reaching the same conclusion is still worth noting.
Where the money goes instead
In practice the shift is modest. Take part of the budget planned for AI-focused site rewrites and move it to three things: earned coverage with the outlets and journalists your category’s answers already cite, a review collection programme with real customers, and creator work on the platforms that show up in the citations, which for cars meant YouTube.
The website keeps its job. Hold the owned work to a standard of accuracy, and spend the growth budget where the answers come from. We have argued before that SEO vendors should report AI Overview coverage weighted by search volume, and the same discipline applies here. Know which sources the answers in your category cite before deciding where to spend. We looked at the risk for established names in Why AI Search Keeps Erasing Familiar Brands.
What I would do this quarter
Write down twenty questions a buyer in your category asks an AI assistant. Run each one in the assistants your buyers use and record every source cited. Sort the sources into three columns: your own site, outlets and creators, and review or community sites. Count each column. If your own site is a small column, as Toyota’s was, that tells you where the next dollar of AI visibility spend belongs. Repeat the count next quarter and compare it with what you spent in each column.
Source: Cision via PR Newswire
