Forrester's latest data shows generative AI answer engines have overtaken vendor websites as the top information source across B2B buying stages. Adobe's response wasn't a new tool — it was an organizational overhaul. Forrester reports that across the Discover, Evaluate, and Commit stages of the B2B buying cycle, generative AI answer engines like ChatGPT and Gemini have become buyers' top-preferred information sources, overtaking vendor websites and product experts. By 2025, 89% of B2B buyers had used generative AI in their purchasing process; by 2026, that number will rise to 94%. This shift matters for demand gen teams because the intermediary has changed sides. Vendor websites and product experts work for the vendor, while answer engines work for the buyer. The same information request now comes with completely different loyalties. If your brand isn't featured in AI-generated answers, you're invisible before a prospect ever reaches your website or talks to your sales team.

Adobe Couldn't Buy a Fix — So It Built One

When AI disrupted how Adobe's buyers learned about its products, Adobe couldn't purchase a solution; no vendor had one. Forrester analysts Joe Cicman, Chuck Gahun, and Phyllis Davidson noted that this led to an adaptation of the operating model rather than a simple technology deployment. Adobe established clear ownership for AI-mediated visibility, installed a continuous review cadence, expanded measurement beyond traditional web analytics, embedded human oversight into AI-assisted workflows, and created mechanisms for sharing learnings across teams. Adobe.com became what Forrester calls a "living laboratory," with lessons feeding back into products like LLM Optimizer. The Forrester analysts were blunt: "I expected a story about AI implementation. Instead, I found a story about operating model adaptation in response to changing buyer behavior." That distinction is crucial.

Why This Is an Ops Problem, Not a Content Problem

Most demand gen teams hearing "AI answer engines are replacing search" instinctively reach for the content lever first: write more, optimize differently, or hire an AEO specialist. However, Adobe's experience suggests that while the content lever is necessary, it is insufficient. The real constraint is organizational. Who owns AI-mediated visibility? What's the review cadence? How do you measure whether you're being cited, summarized, or excluded by ChatGPT and Gemini? These questions often lack clear answers in most B2B marketing organizations because responsibility spans SEO, content, product marketing, web development, and RevOps. Adobe's broader messaging frames AI as infrastructure rather than a campaign. This framing implies process redesign across the GTM function, not a pilot project run by one team. Their multi-model posture, supporting both Gemini Enterprise and ChatGPT Enterprise ecosystems, reinforces this. You can't instrument visibility across answer engines if your measurement and content operations are locked to a single platform.

The Workflow Angle Most Teams Miss

There's a compounding factor that makes this harder to ignore. Gemini is integrated into Google Workspace, and Copilot is part of Microsoft 365. Your buyers aren't opening a separate browser tab for AI-assisted research; they're doing it within the same apps where they write emails and build spreadsheets. This means AI-mediated buying behavior isn't a discrete event you can track with a UTM parameter; it's woven into daily work. A procurement lead asks Gemini to compare three vendors while drafting a requirements document. A VP asks ChatGPT to summarize analyst opinions before a board meeting. If your brand isn't surfaced in those moments, you won't know you lost the deal. There's no bounce rate to measure, no abandoned form to retarget.

What This Means for Your Measurement Stack

Traditional attribution already struggles with B2B's long cycles and multi-threaded buying committees. AI intermediaries complicate this further. The buyer's first meaningful interaction with your brand might occur inside a chatbot you can't monitor. Adobe's response—expanded measurement, continuous review, and human oversight—maps directly onto what marketing ops teams need to build: a governance layer for AI-mediated discovery. This includes monitoring how answer engines reference your brand, testing whether structured content changes affect AI citations, and connecting those signals back to account-level engagement data. This is directional attribution work, not proof-of-causality work. But directional is enough to act on if the alternative is flying blind. The trade-off you're accepting: building this capability requires headcount or agency support, and the ROI won't be clear for quarters. However, the trade-off of not building it is worse. If 94% of your buyers use AI to vet vendors and you lack visibility into how those systems represent your brand, your pipeline forecasts are based on unvalidated assumptions. Adobe didn't wait for someone to publish a playbook; they wrote their own under pressure, with real consequences. Most B2B marketing organizations won't have Adobe's resources, but the operational blueprint—assign ownership, set a review cadence, expand measurement, keep humans in the loop—doesn't require a large budget. It requires acknowledging that the intermediary has changed, and your operating model hasn't caught up yet.