How AI Search Is Reshaping B2B Discovery and Demand

A roundtable conversation with Dave Blanchard, Isaac Voss, and Joshua Strawczynski on why traffic numbers no longer tell the whole story, and what CMOs and CEOs should be measuring instead.

Google AI Overviews can cut organic clicks by 38%, according to a field study covered by Search Engine Journal. At the same time, Demand Gen Report finds that 51% of B2B buyers now use AI chatbots for research, and G2 Research puts GenAI chatbots at 17.1% as the number one source influencing vendor shortlists, ahead of software review sites.

For B2B marketing leaders, the implication is immediate. Discovery is shifting away from website visits and toward AI mediated visibility. That changes what counts as performance, how credibility gets established, and where demand is influenced before a buyer ever reaches your site.

To unpack what that actually means in practice, we brought together three perspectives. Dave Blanchard, a fractional CMO who works with small business owners on growth strategy. Isaac Voss of Rocket Crew, who builds AI ranking systems for service businesses across categories. And Joshua Strawczynski of jMarketing Agency, who specializes in complex B2B purchase decisions across Europe, Australia, and Africa.

The search experience has changed faster than most teams planned for

Dave Blanchard framed the shift this way:

“You used to go on to Google, and you’d type a question or do a search, and you’d get a page of blue links that would go to websites, and the game was to get your links up at the top of the first page. But now we have people not going to Google at all, or if they go to Google, they get a little AI result, and the blue links are farther down the page. So, it’s changed the experience for users and that’s impacted how businesses get found.”

That observation aligns with broader research. Kellogg Insight describes the rise of zero click search, where users get answers from AI summaries instead of visiting publisher websites. Visibility can increase while clicking decline.

Joshua pushed back on treating that shift as a single story:

“The question actually needs to be a lot more nuanced. When we talk about businesses, there’s an enormous difference between a plumber and a high-end consultancy for a Fortune 500 company. Someone looking for a plumber to unblock a toilet, whatever comes up first if they’re available, AI can shortcut things. If you’re looking for a high-end consultancy, we’ve got a lot more psychology involved.”

He also flagged a bigger structural change that most marketing teams have not absorbed yet: “AI now shows up for 51.5% of all searches on Google. It’s taken away 15% of Wikipedia’s traffic. It is important. ChatGPT has just introduced their own pay per click advertising. Whatever we talk about today, I think is strongly going to change in the coming weeks and months.”

Isaac focused on what that compression looks like at the visibility layer:

“I went and looked this morning. If someone’s searching for what we do, the AI only gave me the names of three companies. So, it’s a good thing I showed up as number two. Otherwise, anyone searching for AI may not find me. They may just go, well; these are the three experts in this consultative field, and then they deep dive those three.”

The implication is straightforward. Search did not disappear. It became more compressed, more conversational, and more selective about which brands get surfaced as credible sources.

Traffic still matters, but it no longer tells the whole story

Dave asked the question every CMO is asking right now: “I’ve heard one agency that said they have a client that’s had a 45% decrease in traffic as a result of AI adoption. So ]the next question.is: What’s the impact of that change?

Isaac responded by separating the channel from the playbook:

“I would tell you firstly that organic search is definitely still a channel I would bet on. It’s not just the old version. Traditional SEO as a narrow traffic game, we all know, is gone. But organic visibility as authority, as trust, having a demand capture system is essential. The deeper assets behind organic search have now become more valuable.”

Joshua brought the field data:

“AHREFs just did research recently and found that the top-rated SEO companies have seen a 35% drop in their click through rate. So of course we’re seeing enormous impacts across the board. But what will never go away is the way that people actually make their final determination.”

The practical shift is straightforward. CMOs can no longer rely on sessions alone to understand performance. They need a wider view that includes whether the brand is being cited in AI answers, mentioned in chatbot conversations, or entering the shortlist earlier in the buying process.

The new standard is presence, not just visits

Dave named the measurement question: “Instead of measuring traffic, I measure the presence. Okay, so how do I measure presence? I know how to measure traffic. AI presence, how do I measure that?”

Isaac walked through how he thinks about it:

“You have to regularly test how ChatGPT, Gemini, Perplexity, Google AI overviews describe your category, and then ask buyer intent questions, not vanity questions, and see if you’re surfacing. The visibility layer into AI ranking is difficult because it’s not about keywords. It’s about hundreds of actual human questions and ways that people in different neighborhoods ask about different things.”

Joshua argued the answer is more humanistic than it sounds:

“Where can you see yourself being mentioned by truly reputable sources? Where has a newspaper talked about you? Where has your local municipality mentioned to you? The simplest one is go into Google and see if those reputable people are writing about you. Not some crappy blog, not some weird Facebook accounts, but real media that really matters in your area. If you’ve got stuff that’s recent that talks about you, you’re doing a great job.”

Isaac agreed that the signal layer is broader than most teams realize:

“You have to have all your signals distributed across, I would say, around 20 different platforms. From citations to social, those testimonies and those real human stories and the real community connections. And you have to have backlink authority. You really have to have hundreds of real high-quality domain authority backlinks coming from those local institutions or whatever the trust sources are in your industry.”

A practical presence score for a B2B marketing team can include several indicators.

Frequency of brand mentions in AI generated answers for priority buying questions.
Citation rate in AI summaries tied to category, solution, and competitor terms.
Share of voice for stakeholder specific queries across search and chatbot environments.
Branded search lift and direct traffic changes that follow AI visibility.
Sales team feedback on whether prospects mention AI tools during early conversations.

Buyer questions need to be answered by role and by decision stage

Dave asked Joshua to unpack the time dimension of complex purchases: “Is there a time dimension for this? Do we answer different questions at different times?

Joshua’s answer was rooted in a number every B2B marketing leader should know: “Gartner’s extremely clear on this. People want to complete 57% of their purchase decision cycle before they’re willing to speak to a salesperson. But the website won’t tell me the damn price. So how can I complete that 57% before I speak to a salesperson?

He then named what buyers are looking for when AI delivers them a shortlist:

“Even though AI has told me these are the three best, that doesn’t win you the client. What wins you the client is risk mitigation that this is the correct choice for me and I’m not going to get burnt. And the higher that product is, the more expensive it is, or in B2B the more chance it’s going to cost you your job if you mess it up, the more you have to pay attention to this.”

His framework includes nine signals’ buyers look for: Product performance, expertise, product durability, pricing transparency, production capacity, timeliness, similar clients, customer service, and return policy or guarantee.

“When you answer all nine of those questions in one place without needing to click around, the chance of someone inquiring increases between 40 to 60 percent. If you ask questions about that person so you can give them advice specific to them, and this is what’s relevant to AI, then the chance of them inquiring increases to 84%.”

Isaac connected that directly to AI behavior:

“What we’re talking about firstly is the big shift is recognizing the current and the future is measuring presence, not traffic. Companies just need to ask, are we showing up in AI answers? Are we cited? Are we summarized correctly? Are competitors recommended instead of us? What questions do customers ask before they buy? Do we appear for those questions? Do AI tools understand our category, our location, our offer? And do they believe in them? Are they from reputable sources?”

Build content paths for the buying committee

Dave teased out the structural implication: “It’s conceptually a landing page for the admin, a landing page for the tech guy, a landing page for the HR person, a landing page for the funder, that answers all their questions in one place.”

Joshua confirmed the principle with a B2B case study:

“In B2B, there’s usually between five to seven different people in the purchase cycle. We had an interesting case study recently with a website we made for a medical software business out of Finland. There were five distinct buying groups involved. The doctors wanted the software because it sped things up. The administration of the hospital. The hospital board who ultimately signed off on it. The tech people. And the NHS who would eventually fund it.
All five of those groups have different fundamental questions. So really effective marketing is understanding what the questions are for all five of those groups and being able to provide them in one place for that group.” And he tied it back to the AI layer: “Why is that a smart thing for AI? Well, AI is highly personalized. It knows that you’re the tech guy. And so, it’s going to go crawling through each of those pages, find the one that’s relevant to you, and answer the questions which it knows you need answered. So, until you have those foundations, AI can’t do what AI is really good at doing.”

A strong content structure for any complex B2B sale should include a role specific overview written in plain business language, technical depth where technical evaluators expect it, risk, compliance, or budget justification for financial stakeholders, and clear internal sharing paths so one stakeholder can send another to the right page in seconds.

AI also helps you find buyers, not just influence inbound search

Dave introduced the inverse use case: “I’ve been playing around with AI to find signals for buyers for outreach. It’s really good at that sort of research and identifying signals and ranking prospects, so that I can find the people that are in that 4% buying window. I want to be in front of them, not the rest.”

Joshua’s 4% reference comes from a sobering truth about marketing budgets:

“4% of people are in the purchase cycle at any given time. If I’m a CMO with a finite budget, there’s two things I can do. I can go and try to reach out to more people, knowing that 96% of what I spend my money on is going to people that are not ready to purchase. Or I can focus my attention on the 4% that are and question why they didn’t move forward.”

Isaac added the operational lens:

“When I think about buyer intent signals, I think about, do I have evidence they visited my website? Did they visit my competitors? Did they look up at my competitor? Were they researching the subject matter? If only one out of 25 people that I’m possibly marketing to, want to talk with me right now, then I must think carefully. Do I want to throw mud against the wall, or do I just want to talk to the people who have given me enough signals to say, yes, I’m interested?”

The picture that emerges is two sided. Inbound visibility on one side, outbound precision on the other. Better targeting without better answers will not convert. Better answers without better targeting will move too slowly.

Strong foundations still win in an AI mediated market

Toward the end of the conversation, Dave summarized what he was hearing: “Once you have those foundations, AI does its thing and you don’t have to worry so much about it.”

Joshua agreed, but reframed it as a discipline problem more than a technology problem:

“What will never change is the fundamentals that humans are looking for when they make these decisions. The better that we can work on those foundations and present them, the more we feed the LLMs what they need to make a really honest and valuable decision for us.”

Isaac brought it back to leadership behavior:

“The companies that win over the next 18 months are going to be the ones that make it easier for humans and machines to trust. The ones that are cleaning up their digital presence are strengthening their reviews. From clarifying their offer to publishing better proof, then monitoring how AI is representing them and connecting that whole discovery process to fast follow up and revenue operations. More trust, more clarity, more qualified demand. That’s really the opportunity that’s presented before us.”
His closing advice for CEOs and CMOs was direct:
“Don’t treat this as a panic moment. Don’t treat it as a minor SEO update. It’s a strategic visibility shift, and it’s a season to be the company who embraces and puts these things in place. The ones that are sitting around going, I’m losing all my clicks, my SEO’s changed, what do I do? Those aren’t the answers. It’s these fundamentals we’ve been talking about today.”

Joshua’s parting recommendation was aimed squarely at marketing managers being handed a top-down mandate to figure AI out:

“If you’ve been given that mandate, my recommendation would be, do it, but create a testable model where you’re allocating a certain amount of budget, a certain amount of time, and a very clear way of evaluating if it’s working. And don’t fall into the tricks of SEO from 15, 20 years ago, where an agency goes, look; you’re showing up for these 10 or 20 or 30 terms. Isn’t that fantastic? Because if no one was searching for those terms, if you can’t physically see a change to your bottom line, what’s the opportunity cost of what you could have done with that time, money, and concentration?

The bottom line

The conclusion is simple. B2B discovery is shifting from click capture to answer visibility. Traffic still matters, but it no longer explains whether marketing is shaping demand in a world where buyers start with chatbots, AI summaries, and role specific research paths.

The teams that adapt fastest will measure presence alongside visits, build content for the full buying committee, and use AI on both sides of the funnel to surface better answers and better prospects.

If you are reassessing search, content, and outreach for this new environment, start with one step: audit whether your brand appears in AI results for the buying questions that matter most. That is where the next visibility gap usually shows up.