7 September 2026 · 5 min read
The figures below come from AI Visibility Reports Avid ran during 2026. Brand names, categories, competitors and markets have been removed. These are diagnostic reads, not campaign results.
In one read, a brand's own website was the single most retrieved source across the entire study. More than any publisher, more than any competitor, more than any comparison site.
On the questions where a buyer is still choosing what to buy, that same website ranked 12th and 7th, and accounted for roughly 7% of the third-party editorial total.
Both things are true at once, and the tension between them is the whole point.
The easy line is that AI has replaced your website. It has not, and saying so to a client is both false and unhelpful.
What has actually happened is narrower and more useful to know. Your own pages answer the people who already know your name. They get retrieved constantly for how the product works, what it costs, how to get support. They are close to absent at the moment someone is still deciding what to buy, because at that moment the model is not looking for you. It is looking for someone impartial.
Owned content is not the problem, and it is not the fix. It is doing a job, just not that job.
The next instinct is to find which sources the engines cite most and go there. That breaks down quickly.
In one read, independent editorial was the largest kind of source in one market, at 34.9% of citations. In the other market the largest was retailers and operators at 34.6%, with editorial behind them at 27.9%.
Nobody can place a piece into a retailer's own website. In that second market the biggest kind of source was not a route to anything. It could be measured, and it could not be bought.
Which sources are buyable and which are merely large changes by market and by question. That is a decision to make against the specific answers you want to move, not against a category average.
Splitting one read's questions by how the brand is doing on each gives a shape that keeps repeating:
The instinct is to fight where you are contested. But the two ends of that list need opposite work. Where the brand already holds the answer, the sources deciding it are institutional: reference pages, official listings, the places quoted because they are authoritative rather than because they are interesting. Where the brand is missing, independent editorial decides 21 of those 45.
A single plan of commissioning editorial across all of it would have been wrong, on the brand's own data, for a third of the questions.
And on the questions where someone describes a problem without naming anyone, most answers name no brand at all. That is not a fight being lost. It is ground nobody has claimed yet.
Read the source mix of the specific answers you want to move, rather than the average across your category. Ask which of those sources can actually be commissioned. Then start where the ground is unclaimed rather than where the fight is loudest, because an unclaimed answer is cheaper to win and worth exactly as much.
10 September 2026 · 6 min read
Before any engine can quote your title, the page has to clear conditions that have nothing to do with the writing. The six that decide it, the two halves of a quotable page, and the crawler distinction most robots.txt files have never made.
10 September 2026 · 6 min read
OpenAI's own documentation says a sponsored placement sits below the answer and cannot change it. That makes the ad and the recommendation two different purchases, out of two different budgets.
9 September 2026 · 6 min read
The largest kind of source in an AI answer is often one nobody can buy. Here is the subtraction that shows what is actually left to commission in your category.
See how your brand appears in AI answers, and where trusted publisher content can strengthen its visibility.
