Demand Generation in the Age of AI Search: Six Perspectives, One Problem, and What to Do About It

← View all articles

In 2026, 68% of Google searches end with no click at all. And that changes everything about how your B2B company should think about demand generation.

Six organizations across different disciplines have published research on what this means. They didn’t reach the same conclusions. But together they describe the same problem better than any one of them alone. And the problem has direct implications for any B2B company that relies on content to attract and convert customers.

Why Traffic Stopped Being the Core Metric of Demand Generation

For years, B2B demand generation logic had a simple anchor: organic traffic. More visits to the site meant more leads, more pipeline, more revenue. SEO existed to bring people to the site. Content existed to convert them.

That logic is being dismantled in real time.

In an analysis published by SparkToro in 2026, 68.01% of Google searches ended with no click at all in the first four months of the year, up from 60.45% in 2024. Rand Fishkin attributes much of that acceleration to AI Overviews, present in more than 20% of searches and responsible for a nearly 60% drop in click-through rate when they appear.

My read: that number is going to keep climbing. Google has every incentive to keep the user inside its own ecosystem, and the AI Overview is the most efficient mechanism it has ever built to do that. Betting on the trend reversing is a low-probability bet.

At the same time, the volume of questions asked directly to generative AIs like ChatGPT, Perplexity, Claude, Gemini and Copilot is growing sharply. The B2B decision-maker who used to open 10 comparison tabs now asks an AI and gets back a response with 3 to 5 cited brands, context on each one, and in some cases a direct link to a demo.

And unlike Google, where you can work to climb the ranking, in generative AI responses the logic is different: you need to be recognized as a reference before you get cited, and that recognition is built cumulatively, over time, from content the AIs deemed trustworthy enough to reproduce.

This connects to what I explored in detail in the article how to improve brand visibility in AI responses: SEO, GEO and AEO aren’t competing strategies, they’re layers of the same visibility system that now needs to work across multiple engines at once.

Six Perspectives on the Same Problem

Search Engine Land recently published an analysis using the blind men and the elephant metaphor to describe the current state of the discussion around demand generation in AI search. Six organizations across different disciplines — SEO, public relations, analyst relations and media measurement — are touching different parts of the same animal and reaching conclusions that look contradictory but are, in fact, complementary.

It’s worth understanding what each perspective captures, and what it conveniently misses because of the angle it’s positioned at.

Perspective 1: SparkToro and Audience Attention

SparkToro’s perspective, anchored in search behavior data, focuses on where your ICP’s attention actually is. Fishkin’s recommendations are pragmatic: replace traffic with a correlation dashboard that tracks brand and demand signals over time, invest in channels you don’t control without obsessing over traffic back to the site, and keep publishing on-site content because it still influences AI Overviews even when it doesn’t generate clicks.

What SparkToro sees well: the shift in where attention is. What it doesn’t see as well: the specific mechanism by which that attention translates into pipeline in long, complex B2B sales cycles.

Perspective 2: Fractl and the Tactical GEO Hierarchy

Fractl’s research with Search Engine Land, presented at SMX Advanced in 2026, surfaces two data points that deserve special attention.

The first: in 2025, 82% of consumers found AI search more useful than traditional search. In 2026, that number dropped to 54%, a 28-percentage-point decline in a single year. Trust in AI as an answer source is in free fall before most companies have even structured any GEO strategy.

My take on that data point: it’s both concerning and opportune. Concerning because it suggests the channel is maturing faster than the strategies meant to feed it. Opportune because companies that build credibility now, while the market is still adjusting, will have a structural advantage once trust stabilizes at a new baseline.

The second relevant data point is the tactical GEO hierarchy Fractl proposes: FAQ optimization has 49% adoption and is considered high-risk because it’s easily replicated. Topical authority, brand mentions and structured data are considered table stakes. What creates real advantage is original data, proprietary research and digital PR — the type of content AIs need but can’t generate on their own.

This aligns with what academic research like the GEO study by Aggarwal et al. (Columbia University, 2023) had already identified: content with external citations, verifiable statistics and authoritative language has significantly more presence in AI responses than generic, well-optimized content.

The correlation between brand mentions and YouTube impressions with AI visibility sits between 0.50 and 0.74 in Fractl’s research. Backlink counts and ad spend sit below 0.30. That data point, on its own, should change the conversation about B2B marketing budget allocation.

Perspective 3: AMEC and Upstream Measurement

AMEC, the organization behind the Barcelona Principles that have structured PR measurement for more than a decade, published seven GEO Principles and a Practitioner Guide to GEO Measurement in May 2026.

AMEC’s central contribution is organizing GEO measurement into three domains: upstream reputation, which is the earned, shared and owned content that AI models learn from; search and content readiness, which is whether that information is structured and discoverable; and AI output tracking, which is what stakeholders actually see in terms of presence, framing, citations and accuracy.

What AMEC makes clear that many marketers still haven’t grasped: appearing in an AI Overview is an output. Whether that appearance moved someone toward a purchase decision is an outcome. These are different metrics that require different evidence. AMEC’s guide is honest that connecting AI visibility to pipeline requires “combined evidence,” a triangulated set of signals, not a single number on a dashboard.

Perspective 4: Burson and the Credibility Paradox

Burson, one of the world’s largest PR agencies, ran an analysis of thousands of reputation-related prompts across seven AI platforms, covering 85 companies in 10 sectors.

The central finding is what Burson calls the Credibility Paradox: a brand can be cited by an AI and still lose the reputation opportunity if the public doesn’t believe what the AI says about it. Being seen is not the same as being believed.

The data point most applicable to B2B demand generation is the proof-versus-posture split. Attributes backed by observable evidence, such as innovation, products and workplace culture, performed roughly twice as well in terms of credibility as attributes that rely on institutional self-description, such as leadership, governance and corporate citizenship.

AIs are more willing to endorse what your product does than what your leadership says about its own values. That is a direct signal about content prioritization.

Perspective 5: Analyst Influence on B2B Buying Cycles

The perspective most relevant to B2B companies comes from Jamin Spitzer, former head of communications insights at Microsoft, who argues that GEO belongs on the analyst relations desk, not on the SEO team.

The argument is direct: when a B2B buyer asks an AI who leads a category or what their evaluation shortlist should be, the answer is frequently a synthesis of analyst content — Gartner, Forrester, IDC, independent analysts — because that content is exactly the kind of taxonomy-rich, comparative, authoritative material generative engines are trained to reach for.

For B2B demand generation specifically, this suggests that the upstream content that matters most isn’t earned media or product pages. It’s analyst relations and the content those relationships produce.

This is the perspective the other four largely ignore. AMEC, Burson and Fractl gravitate toward consumer-facing signals or brand reputation. None of them address the specific mechanism Spitzer describes: a multi-month enterprise sales cycle in which an AI-generated “consideration set,” built partly from analyst reports, can shape outcomes before the buyer ever opens a Magic Quadrant.

Perspective 6: The Credibility Gap in the Sources AI Cites

The final perspective, based on research by Angela Dwyer at Full Intel, examines which news sources AI platforms cite most frequently and found a gap between citation frequency and the outlets audiences consider most trustworthy.

It’s the mirror image of Burson’s Credibility Paradox, but on the publisher side: just as a brand can be visible but not believed, a publication can be heavily cited by AIs while its own readers consider it less trustworthy than less-cited competitors.

For demand generation, it’s a reminder that getting a mention in a high-citation-volume outlet is no guarantee of credibility transfer. The perceived quality of the source matters as much as citation frequency.

What This Changes in Practice for B2B Companies

Taken together, these six perspectives point to a shift that goes beyond adjusting your SEO strategy.

The question most B2B companies still aren’t asking systematically is: what do AIs know about our company, where did that information come from, and what does our ICP see when it asks an AI about our category?

That question has three distinct components, and each requires a different answer.

The first is visibility: does the company appear in AI responses relevant to your market? This is measurable today, manually, by testing the questions your ICP actually asks in ChatGPT, Perplexity, Claude and Gemini.

The second is framing: when the company appears, with what context and positioning is it described? Is the AI using the company’s current positioning, or an outdated version of how it described itself two years ago?

The third is credibility: is what the AI says about the company the kind of thing a B2B buyer will believe? Is it grounded in product and outcome evidence, or generic mission-and-values self-description?

Together, these three questions are the minimum diagnostic any B2B company should run before making any content and marketing budget allocation decision.

How a B2B Company Disappears From AI Search Without Realizing It

A B2B tech company with five years in the market, a product customers vouch for, and a high NPS decides to understand why demand generation through organic content isn’t growing in proportion to investment.

The SEO report looks reasonable: domain with established authority, indexed articles, some keywords ranking on page one. But organic traffic is stagnant and the leads coming in through content are below expected quality.

The team tests the major AIs with questions about the category. Competitors show up cited frequently, with specific context about use cases and differentiators. The company itself appears in fewer than 15% of responses, almost always as an item in a generic list, without the context that would make the citation useful to a buyer in evaluation.

The diagnosis is clear when read through the six perspectives: the company’s blog has content about features and product launches, useful for current customers, but it doesn’t answer the questions decision-makers ask before buying. The company has no relationships with industry analysts. Published case studies are generic, without specific data AIs could reproduce as evidence. And the positioning across its online materials still uses the vocabulary from three years ago, before a strategic repositioning that was never translated into published content.

Every element the six perspectives identify as upstream for AI visibility — topical authority, original data, mentions in trustworthy sources, analyst relations, verifiable product evidence — is either missing or outdated.

The fix isn’t publishing more. It’s publishing differently, with judgment about what AIs need in order to cite a brand credibly.

How AVPIA Helps B2B Companies Build Presence That AI Recognizes

The link between AI visibility and B2B demand generation isn’t just a marketing concern. It has a direct impact on the quality of the lead that reaches the sales team.

The B2B buyer who was recommended by an AI arrives with context about the brand, a prior perception of credibility, and a different disposition for the first conversation. They don’t need to be convinced the problem the company solves is real. They’ve already gone through that stage before arriving.

The AVPIA Virtual SDR receives that lead at a different stage than a lead that arrived with no prior context. Qualification is faster because the lead arrives informed. The conversation goes deeper from the first contact. And the sales cycle tends to be shorter because it starts from a foundation of trust that the content built before the commercial approach.

For B2B companies building this visibility strategy, the AVPIA Platform provides the prospecting and qualification infrastructure that captures the leads generated by organic presence and works them with the same consistency the content built. It makes no sense to invest in GEO and AEO to generate awareness if the sales operation can’t convert that awareness into pipeline with speed and judgment.

Want to understand how ready your sales operation is to receive leads coming in through AI search? Schedule a demo and see how the platform works with the lead profile your content strategy is generating.

Why Multidimensional Demand Generation Changes the ROI Conversation

The most practical implication of these six perspectives together is that the success metric for demand generation needs to be multidimensional. Not because it’s sophisticated to have many metrics, but because the problem we’re trying to measure is genuinely multidimensional.

Traffic remains relevant for high-intent transactional searches. Brand mentions and presence in AI Overviews are relevant for top-of-funnel demand generation. Credibility in what AIs say about the company is relevant to visitor-to-lead conversion rate. Analyst relations are relevant to the consideration set of enterprise buyers in long cycles.

None of these dimensions replaces the others. Each covers a different stage of the buying journey.

Fractl’s research found that buyers check an average of 2.4 platforms before validating a purchase decision. That means a company present on only one of those platforms with credibility is losing influence at other stages of the journey that happen before the lead ever identifies itself.

For B2B marketing leaders who need to justify content investment without traffic as proof of results, the framework that emerges from these six perspectives offers a more honest answer than any sessions-and-pageviews dashboard. What you’re building is presence in the sources decision-makers consult before raising their hand. Measuring that requires different metrics than what Google Analytics offers by default.

This theme connects to what we analyzed in B2B demand generation strategy in 2025: lead attraction strategies are evolving from direct traffic to distributed influence, and companies that recognize this shift ahead of competitors build an advantage that takes years to replicate.

Final Reflection

Six organizations touching different parts of the same elephant reached conclusions that look like they compete but actually complement each other. Each captures a real dimension of how AI is changing demand generation. None captures the whole picture.

What this means practically for a B2B company: you need more than one perspective to understand the problem, and more than one metric to know whether you’re making progress.

Traffic still matters where search behavior still generates clicks, especially for transactional and branded searches. Topical authority and original data matter for being cited by AIs with credibility. Analyst relations matter for appearing in the enterprise buyer’s consideration set. And the credibility of what AIs say about your company matters more than simply appearing.

The window to build that presence cumulatively, before competition for space in AI responses gets as fierce as competition for Google ranking, is still open. For how long, none of the six perspectives answers precisely. But the direction is clear.