Forrester's 2026 Buyers' Journey Survey found that the share of B2B buyers using generative AI in their purchase process rose from 89% to 94% in a year. The more consequential finding is the second one: twice as many buyers named generative AI or conversational search as their most meaningful research source than named any other source, ahead of vendor websites, product experts, and sales representatives.
That is not an incremental channel shift. It is a change in where the first impression of your product gets formed, and for most companies it now happens somewhere they have no visibility into and no direct control over.
What actually changed in the funnel
The traditional model assumed a buyer becomes aware of a problem, searches, lands on vendor sites, compares, and shortlists. Every measurement system built in the last fifteen years assumes some version of that sequence.
The current sequence increasingly starts with a buyer describing their problem to an assistant in natural language and receiving a synthesised answer that names two or three approaches and sometimes two or three vendors. By the time they arrive at your site, the shortlist already exists. You are either on it or you are not, and the decision was made somewhere you cannot see.
The traffic implication is not the main story
Traditional search volume is projected to decline around 25% by 2026, with some forecasts putting organic traffic down 50% by 2028. AI referral traffic remains small, around 1% of web traffic, but it is growing quickly and converting well above traditional organic.
“By the time a buyer reaches your site, the shortlist already exists. The decision was made somewhere you cannot see.”
The temptation is to read this as a traffic problem and respond with more content. That misreads it. The problem is not volume. It is that the buyer arriving from an AI answer has already formed a view, and the visit is confirmatory rather than exploratory. A page written to persuade a cold visitor is answering questions that were already answered elsewhere.
What this means for a small venture
By the time a buyer reaches your site, the shortlist already exists.
For a solo founder or small team, this is a genuinely favourable change, which is not the usual conclusion. Being cited in a synthesised answer does not require domain authority accumulated over a decade. It requires being the clearest and most specific source on a narrow question.
Turn what you know into what you own.
Vibepreneur builds structured ventures from professional expertise, with positioning, launch assets, and growth systems included.
Join the WaitlistA large competitor writing broadly about a category is easier to overlook than a small vendor writing precisely about one workflow. Specificity is the ranking factor that matters here, and specificity is the one thing a narrow product has in abundance.
The practical adjustments
Write to answer the question rather than to rank for the phrase. If a buyer asks an assistant how to reduce changeover losses in a mixed manufacturing line, the content that gets cited is the one that answers that question directly, with numbers, in the first two paragraphs.
Put the specifics near the top. A synthesised answer draws from the part of the page that most directly addresses the query, not from your carefully sequenced narrative. Front-load the useful part and put the persuasion below it.
Include your own data. First-party numbers from your own customers or your own operating experience are the material a model cannot get anywhere else, which makes them disproportionately likely to be surfaced and attributed.
What to stop doing
Stop writing content that summarises what is already widely published. That was always low value and it is now actively pointless, because a model can produce it instantly and will never cite you for it.
The next post in this series covers generative engine optimisation for solo founders in practical detail. If you are rebuilding your positioning around this, the positioning studio is the structured version.