The headlines concerning AI have been focused on traffic: When buyers use AI instead of search, they are one-tenth as likely to click through to your website, and with business buyers rapidly adopting AI, B2B companies are feeling the impact in the form of traffic declines of between 10–40% over the past year. That is a huge disruption to a key metric for many companies, so it creates urgency, but the decline in traffic tells only half the story.
Private AI Is Driving AI Adoption In B2B
Business buyers are not only adopting AI more rapidly than consumers —they’re also using it differently. More than half of business buyers report that they are using private AI tools provided by their company. This corporate sponsorship is responsible for making Microsoft Copilot the most widely used AI tool, with 68% of business buyers in Forrester’s Buyers’ Journey Survey, 2025 reporting using Copilot and more than half of them (36%) using a private instance behind their corporate firewall.
AI Is More Than Search
AI may displace traditional search, but it’s not only a replacement of search. Buyers use AI in a wide range of use cases. Buyers now use AI for tasks they once completed by searching, such as researching product information (54%) or making product comparisons (55%). But they also use AI for tasks behind their firewall that are very different, like analyzing RFP responses (48%) or building a business case (47%). Soon we’ll see buyers using AI in more agentic modes: They will use procurement agents to look at vendor meeting transcripts, demos and downloads, and internal debriefs, and they will filter out personal preferences and identify pros and cons, certainties, and discrepancies across the vendors being considered.
Visibility Is Only Half The Story
Providers are right to start with answer engine visibility. That’s the first half of the story. But as business buyers continue to use AI, providers must cater to these more sophisticated use cases. Companies will need to provide buyers with the content, tools, and experiences needed to validate or correct their AI-sourced knowledge, and they will need to explore new models like Model Context Protocol for providing buyers with trusted information that they can use to create context and substance for the internal AI tools supporting buying processes such as analysis, risk management, budgeting, negotiation, and contracting.
