Skip to content
Holmby Lane

Insights·Opinion

The B2B Buying Journey Now Starts in a Chat Window

The anonymous research phase of B2B buying has moved into AI assistants, which compresses shortlists before any vendor knows the deal exists. What that does to pipeline math.

June 8, 2026 · 6 min read · Holmby Lane Research

The B2B Buying Journey Now Starts in a Chat Window

B2B buying always had a hidden phase: the months a buying committee spends researching before any vendor knows the deal exists. Analysts have documented for years that buyers complete a majority of the journey before first contact, and that they prefer it that way. The change is where that hidden phase now happens. Increasingly, it happens in a chat window, and the output is a shortlist assembled by a machine.

What actually changed

The old anonymous phase left footprints: search sessions across many queries, a dozen tabs of vendor sites, content downloads behind forms. Buyers now compress much of that into conversations with an assistant: describe our situation, list the credible options, compare the top three, draft the requirements document. The assistant does the tab-opening, and the buyer reads a synthesis.

Three consequences for vendors:

  • The shortlist forms earlier and shorter. When an engine names three vendors and describes them plausibly, many committees start there. Being the fourth-best-known option used to mean losing sometimes; now it can mean never being evaluated.
  • Your first impression is written by a machine. The engine's one-sentence characterization of your firm (assembled from your entity, your reviews, your coverage) is your first sales conversation, and no one from your team is in the room.
  • The demand shows up disguised. By the time a buyer hits your site, they arrive with formed opinions via Direct or branded search, which is why AI referral instrumentation and self-reported attribution suddenly matter: they are the only visible edges of the hidden phase.

What it does to pipeline math

Marketing models built on intercepting the research phase (rank for the query, gate the whitepaper, nurture the lead) lose surface area as research moves somewhere ads and forms cannot follow. The compensating asset is answer presence: share of voice inside the conversations where shortlists form. That asset behaves differently from traffic. It does not convert or retarget. It preconditions: deals arrive fewer in number, better qualified, already framed by whatever the engine said about you.

The stance to take

Treat the engines as the new first meeting. Know what they say about you (run the buyer prompts weekly; the prompt universe method formalizes this), fix what they get wrong, and invest in the sources they lean on for your category: review platforms, comparison content, credible coverage. And keep perspective: engines do not close deals or replace relationships. They decide who gets the chance to build one, which is exactly why they deserve a line in the pipeline model rather than a paragraph in the innovation deck.

Put this to work

Holmby Lane runs AEO-led growth programs: entity work, citation campaigns, and the content AI engines actually retrieve, measured against your buyer prompts daily.

Keep reading