Does my company show up when buyers ask AI for an outsourcing shortlist?
Probably not, and that is the point. Our analysis of the prompts buying committees use shows the same handful of mega-providers in eight to ten of every ten AI-generated answers. Inclusion is driven by signals that do not require scale.
What the answers actually look like
We analyzed the prompts real buying committees use when they ask an AI model to build an outsourcing shortlist. The same handful of mega-providers appear in eight to ten of every ten answers.
That concentration is worse than search ever was. A search results page gave you ten links and let the buyer decide. A model gives one answer with three to five names in it, and the names below the fold do not exist.
Five things drive inclusion
Publisher list placement. Analyst citations. Named AI products. Content depth. Review density.
Look at that list again. None of them require scale. They require clarity about what you are, and a deliberate plan to build the evidence.
This is why the concentration is not permanent. The mega-providers dominate because they have accumulated these signals over years, not because the models prefer large companies. A specialist with published proof in a narrow category can outrank a generalist inside that category.
Why your website alone will not fix it
Two different things determine whether you appear. What a model can learn from your own pages, and what a model can learn about you from everyone else.
You control the first completely. Clear positioning, structured data, content written so a machine can extract an answer rather than infer one. Necessary, and not sufficient.
The second is corroboration, and it is the harder half. Being named in other people’s content, cited by analysts, placed on published lists, reviewed by real customers. No amount of work on your own site substitutes for it, which is worth saying plainly because a lot of AI visibility advice implies otherwise.
The clarity problem underneath
When a model has to describe you, it works from what it can read. If your own description is hedged, broad, and interchangeable with four competitors, there is nothing distinct to repeat.
The companies showing up made it easy for a machine to understand exactly who they are. That is a clarity problem before it is a technical one, and it is why the work usually starts with positioning rather than with markup.
How to find out where you actually stand
Guessing is expensive here, because the intuition is usually wrong in both directions. Companies that assume they are invisible sometimes surface well in narrow categories. Companies that assume they are known are often absent from the exact prompt their buyers use.
The Discoverability Baseline answers it directly. We test how your company surfaces when buyers ask AI models to build a shortlist, benchmark you against three competitors you approve, and come back in two business days with the evidence, the findings, and the first moves that change it. Every report is personally reviewed before it ships.
Your first one is complimentary. What comes after it is the paid part: quarterly tracking that re-runs the identical panel and shows whether you moved.
One boundary worth stating
The Baseline is a commercial product. It is separate from The Pivot Path, the editorial rankings we publish, including The IPO 25 and the Trust & Safety 10. Buying it does not affect inclusion or position on any list. No company pays to be on those, and no company can.
Get the answer for your own company
How you surface against three competitors you approve, with evidence and first moves. Two business days.