AI search is shrinking traffic, but improving lead quality

The wrong way to read a traffic drop is as a simple loss. In B2B, AI search is pushing more of the research phase up into the results page, which means fewer casual clicks can still produce better buyers. The useful question is no longer how many visits landed, but whether the visits that remain are better qualified, move faster through sales, and convert at a higher rate.

AI search is changing what “visibility” means

Kerry Cunningham’s framing makes sense because it matches what buyers are doing now. AI Overviews and other answer engines are compressing the top of the funnel into a single screen: summaries, comparisons, vendor shortlists, and references pulled from reviews, case studies, and analyst coverage. By the time a prospect clicks through, they have often already done the skimming and the early elimination that used to happen on your site.

That changes the job of B2B content. The goal is not just to catch broad informational traffic anymore. It is to become the answer that AI systems cite when a buyer is trying to decide who belongs on the shortlist. If your content is only optimized for raw clicks, you may be fighting yesterday’s game.

The Seer data shows why citation matters

Seer Interactive’s 2026 AIO study gives the cleanest proof that AI visibility is not all the same. Brands that appeared on AI Overview-present search results pages but were not cited inside the overview saw organic click-through rate fall 67% over 2025. Cited brands, by contrast, earned 120% more organic clicks per impression than uncited competitors.

That gap matters because it shows the buyer is not just seeing your name. They are seeing your name in a trusted, summarizing context. Seer’s own April 24, 2026 update widened the lens further, tracking 53 brands, 5.47 million queries, and 2.43 billion organic impressions from January 2025 through February 2026. It also reported that organic CTR on AI Overview-present queries had rebounded 85% in two months by early 2026.

The practical takeaway is simple: citation changes behavior. If you show up in an AI answer, you are not merely getting exposure, you are influencing whether the searcher bothers to click at all. For B2B teams, that means the best traffic may be the smaller slice that already has intent, context, and trust built in.

Why fewer visits can still mean better pipeline

This is where a lot of marketers get trapped in vanity math. A lower session count can look like failure even if the pipeline downstream is healthier. If AI search has stripped out the early-stage tire-kickers, the visitors who do arrive may be the ones who already understand the category, recognize the problem, and are closer to action.

That is the recalibration B2B teams need to make. Instead of celebrating traffic for its own sake, measure whether the remaining sessions create stronger lead quality, faster sales velocity, and better conversion rates. If a page brings in fewer visitors but more of them turn into sales-qualified leads or progress through the funnel faster, the traffic loss is not the real story.

The sponsored premise is worth testing against the back end, not accepting on faith. A smaller pipeline can be healthier only if the pipeline is actually cleaner. If conversion rates do not improve, or if sales velocity stays flat, then AI search is not filtering for quality so much as removing demand you still need to replace.

Forrester says the buyer journey was already happening off-site

Forrester, based in Cambridge, Mass., United States, has been pointing in this direction for a while. Its State Of Business Buying, 2024 report, published on December 4, 2024, says more than 80% of the B2B buying journey now happens without vendor involvement. The same research points to tight budgets, AI’s influence in buying and selling, negative buying experiences, and long purchase cycles as forces making the process even messier.

The more sobering number is this: more than 80% of buyers are dissatisfied with the provider they choose at the end of a purchase process. That tells you the market is not exactly overflowing with confident, well-served decision-makers. It also suggests that a lot of buying happens under pressure, with imperfect information, and with a good deal of regret baked in.

A later Forrester report summary sharpens the point even more. It says 92% of buyers start with at least one vendor in mind, and 41% already have a single preferred vendor before formal evaluation begins. In other words, the “decision” may be made long before the sales conversation starts. By the time your team shows up, the real job is often to confirm, not create, the preference.

What B2B teams should do differently

If AI search is becoming the front door to evaluation, the content mix has to match that reality. The pages most likely to matter are the ones that help answer narrowing questions: comparison pages, category explainers, case studies with specific outcomes, and pages that make it easy for an AI system to summarize your position accurately.

A practical playbook looks like this:

• Build content that can be cited cleanly. Clear positioning, concrete outcomes, named customers where possible, and tight summaries make it easier for answer engines to lift your proof.

• Watch CRM outcomes, not just traffic. Track lead quality, sales velocity, and conversion rate by landing page and by query intent, then compare AI-visible pages with everything else.

• Keep an eye on the “near-final opinion” problem. If buyers already arrive with a vendor in mind, your content has to win trust fast or it will be skipped.

• Treat review sites, analyst coverage, and case-study ecosystems as part of your search strategy. Seer’s findings point to cited sources as the ones that keep earning clicks.

The point is not to chase more traffic at any cost. It is to make the traffic you do get count harder. In a market where buyers are doing more of the work before they ever talk to sales, a well-qualified click is worth more than a bloated visit count.

Google has started giving owners more control

Google’s response shows this debate has moved from theory to product controls. On June 3, 2026, it announced generative AI performance reports in Search Console and a site-level toggle that lets website owners opt out of appearing in AI Overviews, AI Mode, and AI Overviews in Discover. Google says sites that opt out will not receive traffic or impressions from those generative AI features, but they will still appear in standard search results.

That matters because it gives publishers and B2B teams a clearer choice: participate in the AI layer and measure the result, or opt out and accept the tradeoff. Google’s Search Central guidance also frames AI features as a visibility surface that site owners should measure and manage, which is exactly how it should be treated.

For B2B marketers, the lesson is blunt. AI search is not just shrinking traffic; it is filtering intent. The winning teams will be the ones that stop worshipping volume, start tracking pipeline quality, and build content that AI systems trust enough to cite when the buyer is already halfway to a decision.