REVENUE MOTION FRAMEWORK™ · CHAPTER 3
How AI Is Changing the B2B Demand Funnel
The framework holds. The routes to the buyer have changed.
By Roel Haanappel · Published 16 October 2026 · From the second edition of the Revenue Motion Framework™ ebook
The dark funnel is the part of the buying journey you cannot see. AI has made it darker. Not because buyers stopped researching, but because they now research somewhere else.
The questions that used to bring buyers to your website, your blog and your gated ebook are increasingly answered before they ever arrive. Most demand generation programmes were built on one assumption: that buyers educate themselves on vendor websites and leave signals behind while they do. That assumption is breaking down. The buying carries on. The route it takes has changed.
From search engines to answer engines
For twenty years, search engines ranked pages and sent a click. You competed for position, earned a visit and turned the visit into a signal.
AI assistants such as ChatGPT, Perplexity, Claude and Google’s AI answers work differently. They read many sources, form a view and give the buyer an answer. Machines that rank are becoming machines that answer.
The change is already measurable. In G2’s 2026 survey of more than 1,000 B2B software buyers, 51% said they now start their research in an AI chatbot rather than a search engine. And AI is speeding up a pattern that was already there: buyers are nearly 70% of the way through their process before they engage a seller, and 81% have a preferred vendor by the time they do.
If you are not in the answer, you are often not on the list.
Seven shifts in the B2B demand funnel
- The dark funnel got darker. Research that used to leave signals on your website now happens inside AI conversations. The early visits that warned you weeks ahead are disappearing.
- Buyers arrive informed but unknown. They come late, with an opinion already formed, looking for confirmation and a next step rather than education. Expect fewer page views and fewer form fills from the same amount of buying.
- The educational funnel is collapsing. AI answers the top- and middle-of-funnel questions your content was written for. Gated content and MQL volumes shrink, while pipeline that comes through other routes does not.
- Content now serves four audiences. People decide. Search engines rank and send a click. Large language models cite. AI agents act. All four are best served from one structured content source, with the website as one of its outputs rather than the only place your knowledge lives.
- The shortlist is set before sales gets involved. This was a trend before AI. AI made it effortless, so now it is the norm.
- Reviews and peer proof feed the machines. Review sites, communities and peer platforms are among the sources AI answers draw on. The volume and recency of your customer proof now shape how AI describes you.
- Agents are starting to buy. Procurement platforms already use AI agents for supplier research and comparison. For complex B2B technology purchases, agents prepare the comparison and humans still decide, for now. Gartner expects AI agents to intermediate more than $15 trillion of B2B spending by 2028. Treat that as a forecast, not a fact. Either way, agents favour vendors whose information is clear, structured and verifiable.
What most B2B companies get wrong
They publish only on their own website. Good content is still the foundation. But AI answers lean heavily on sources you don’t own: press, analysts, reviews, communities and expert voices. Content that lives only on your site reaches a shrinking share of buyers.
Nobody owns the routes. PR chases news. Analyst relations chases rankings. Communities run on their own. Executives post when they have time. Being the answer falls between all of them.
They keep measuring the old funnel. When MQLs and click-based attribution decline while buying continues, the wrong debate starts: is demand generation dead, should we switch to brand? Demand generation is not dying. It is moving to routes the old metrics cannot see.
They produce generic AI content. As AI-written content floods every channel, it becomes invisible to buyers and to the models alike. What both value is what cannot be generated: original data, practitioner experience and a clear point of view.
How to be the answer, wherever buyers ask
The response is not a new framework. It is a new standard for the one you have. Content still wins; the routes to the buyer have changed. In practice that means three things:
- Choose the topics you want to be the answer for. Not every keyword. The handful of buying questions where your expertise and your customer proof are strongest, and where your buyers are asking AI for advice.
- Get every route working towards them. Marketing, PR, analyst relations, communities, partners and executives each feed the sources AI draws on. Being the answer is a shared outcome with a clear owner, not a website task.
- Measure after the click. Track how often AI answers cite you on your chosen topics, ask new opportunities where they first heard of you, and watch branded search and pipeline velocity. These replace MQL volume as the early signals of demand.
How each framework phase changes, and the open questions on measuring GEO: GEO and the dark funnel: practitioner notes.
What this means for the Revenue Motion Framework™
The Revenue Motion Framework™ still describes how enterprise demand works: align, sense, orchestrate, convert and optimise, as one continuous motion. What changes is what goes into each phase and what comes out. The free ebook covers all five phases, each ending with what changes in the AI era.
Want to know where your company stands in AI answers today? MatterMore’s AI Search & Demand Advisory starts with that baseline. See the advisory →
FAQ
B2B buyers now research inside AI assistants such as ChatGPT, Perplexity and Google’s AI answers instead of on vendor websites. They arrive later, better informed and often with a shortlist already formed. The early website signals demand generation relied on are disappearing, while the buying itself carries on.
No. Demand generation is moving to routes that click-based metrics cannot see. MQL volumes and gated-content downloads fall, while pipeline that comes through AI answers, peers and earned media keeps flowing. The task is to be the answer on those routes and measure them.
The dark funnel is the part of the buying journey a vendor cannot see: research in communities, peer conversations and now AI assistants. AI has made it larger, because research that used to leave signals on your website now happens inside AI conversations.
Choose the few buying questions you want to be the answer for. Publish clear, structured content on them, and make sure press, analysts, reviews and expert voices say the same. AI answers draw heavily on sources you don’t own.
Being visible in AI answers is a shared outcome, not a website task. Marketing, PR, analyst relations, communities and executives all feed the sources AI draws on. Agree who contributes what, and give one person ownership of the result.
Track how often AI answers cite you on your chosen topics, ask new opportunities where they first heard of you, and follow branded search growth and pipeline velocity. Review them weekly alongside pipeline, instead of MQL volume.