Absent on the surface the buyer uses
Meta description: An AI answer may name a much smaller set of providers than a results page — that is the demand meeting companies miss.
Miklos Roth
The demand review in these companies has a quiet assumption running under all its charts: that presence means the company showing up — its site, its content, its brand properties, ranking and converting. Every metric in the deck measures a property the company owns. And in the stories I am telling, the buyer’s journey has moved: the early questions are asked in AI engines, the answers are assembled from sources the engine selects, and the shortlist forms there. The company is absent from that surface. Its owned presence is undamaged and increasingly irrelevant, like an excellent shopfront on a street the buyers no longer walk down.
[AWAITING MIKLOS — scene: a demand review that still equates presence with the company site]
What does the company currently call presence?
Presence, in the dashboards I read, is a set of owned-surface metrics: impressions, rankings, site traffic, brand search volume. These measure the company’s availability to buyers who come looking. The assumption embedded in the dashboard is that buyers come looking — that the journey includes a visit to the company’s surfaces. The AI answer layer interposes itself earlier: buyers arrive pre-informed, their questions answered, their criteria set, sometimes their shortlist closed, without the company’s surfaces being touched. The company has presence on the path it measures. The path moved.
Which buyer question has no company source in the answer?
The surface that matters is the recommendation set itself. A conventional search results page may show many links, and a business in fifth or sixth place can still receive visitors. An AI answer may mention a much smaller set of providers. That is an important commercial distinction: the companies named in the answer are, functionally, the demand meeting. Everyone else was evaluated and set aside without ever knowing the meeting happened.
The diagnostic I run is naive on purpose: take the ten questions a buyer asks before contacting any vendor, and check whose material informs the answers. In the companies I am describing, the company’s material informs few or none — not because the company lacks content, but because its content was built for the owned-surface journey: product pages for visitors, conversion copy for arrivals. The pre-contact questions are different in kind — comparative, situational, skeptical — and the sources that answer them are built for exactly that. The company wrote for its shopfloor. The buyer’s questions are being answered in a building the company has never entered.
Why does a new guide fail to close that specific gap?
Read more: Miklos Roth.
The false explanation is content: publish a guide, cover the topic, be present. I have watched guides published into this gap and disappear, because the gap was never a missing document. The sources the engine selects for pre-contact questions carry properties a vendor’s guide struggles to fake: disinterest, corroboration, a track record of answering without selling. The company’s guide is a sales artifact wearing educational clothing, and the engine’s selection — tuned on what satisfies skeptical questions — routes around it. More content on owned surfaces deepens the owned presence. The gap is on unowned ground.
What pattern marks a commercial visibility gap?
The distinction I use is informational versus commercial absence. An informational gap costs traffic — regrettable, recoverable. A commercial gap costs the frame: the criteria, the category, the shortlist logic — and it costs it before any metric the company tracks can register the loss. The pipeline data arrives late and misattributed: deals lost to “no decision,” competitors entering already favored. I have learned to treat those pipeline anomalies as the downstream signature of an upstream absence. The company was not outbid. It was unmentioned in the room where the bidding rules were written.
I would resist selling presence in that set as a replacement for search work. It is a more complicated continuation of the same underlying work: helping the right customer discover, understand, and choose the right business. A serious report distinguishes a sampled AI mention rate from organic search positions, actual site visits, qualified leads, and completed revenue. These are different quantities, and the meeting’s minutes are written in all of them.
[AWAITING MIKLOS — scene: second scene]
What question names the surface before the asset?
Marketing planning asks what we should publish — an asset question, answered with a calendar. The question that opens this problem asks where the buyer’s decisions now actually form — a surface question — and insists on an answer before any asset is commissioned. Companies that ask it discover their content operation is well aimed at a surface of declining consequence. Companies that skip it keep publishing excellent material into a room the buyer left.
The question that puts the invisible meeting on the agenda: in the answers where this company is absent, which competitors are present — and what evidence have they made accessible that this company has not?
PR excerpt (80–120 words, Miklos’s own words): A conventional search results page may show many links, and a business in fifth or sixth place can still receive visitors. An AI answer may mention a much smaller set of providers. The companies named in the answer are, functionally, the demand meeting — everyone else was evaluated and set aside without knowing the meeting happened. Presence in that set is not a replacement for search work; it is a more complicated continuation of the same underlying work.