HubSpot yesterday published data showing that AI search has become the single strongest predictor of purchase intent among CRM buyers - surpassing product demos, review sites, and sales calls - based on a survey of more than 3,000 decision-makers conducted in January 2026.

The announcement, updated on June 12, 2026, adds a new dimension to the debate around declining organic traffic. For months, marketing teams have grappled with falling website visits. The HubSpot data reframes that conversation: the issue is not simply that buyers are disappearing from the web, but that they are reappearing somewhere else - inside AI platforms - and converting at measurably higher rates when they do.

According to HubSpot, 42% of CRM buyers used AI search during their evaluation process. Those buyers were 36% more likely to purchase than those who did not use AI search at any point. That single figure - a 36-percentage-point lift in purchase probability - is, according to HubSpot's own framing, a larger predictor effect than any other factor the survey measured, including live product demonstrations.

The survey and what it measured

HubSpot surveyed more than 3,000 CRM purchase decision-makers worldwide in January 2026. The objective was to identify which touchpoints and behaviors most strongly correlate with a completed deal. The findings were drawn from HubSpot's own customer base, which spans nearly 300,000 businesses.

The methodology matters for interpreting the results. The survey captures self-reported behavior from buyers who ultimately reached a CRM vendor - meaning it measures what distinguished buyers who purchased from those who did not, within a pool that had already engaged with the category to some degree. It does not measure all-category buyer behavior from scratch. That said, the scale - 3,000 respondents across multiple geographies - makes it one of the larger datasets on AI search behavior in a B2B purchasing context published to date.

AI search was defined, in the context of the survey, as buyers using platforms such as ChatGPT, Claude, and Gemini to ask questions during their evaluation. These are conversational AI tools that generate synthesized responses rather than lists of links. The behavioral distinction from traditional search is significant. A buyer consulting ChatGPT about CRM software has already moved past the keyword-query stage; they are asking structured questions and receiving structured answers that name specific products and frame comparisons. By the time such a buyer reaches a vendor's website - if they reach it at all - they may already have a shortlist.

Organic traffic is down 27% year over year

The purchase-intent finding arrives on top of a traffic finding that HubSpot first reported in April 2026. According to HubSpot's proprietary data, organic traffic to customer websites fell 27% year over year. That figure now appears in the June 12 publication with added context: the decline is not primarily an optimization failure, but a behavioral shift. Buyers have moved their research activity to AI platforms, and that migration is accelerating.

The 27% figure, when placed alongside the January 2026 survey results, produces a specific picture. Fewer buyers are arriving via traditional organic search. But among those who do arrive after using AI search, purchase rates are materially higher. The implication, as HubSpot frames it, is that AI search functions as a pre-qualification layer - buyers who have gone through that process arrive further along in their decision.

This pattern aligns with findings that PPC Land has tracked across multiple research publications throughout 2025 and 2026. A June 2025 analysis found AI search visitors worth 4.4 times more than traditional organic visitors in value-adjusted terms, citing research that documented higher purchase intent and research completion among AI-referred visitors. Separately, data published in November 2025 showed AI traffic converting to sign-ups at 1.66% compared to 0.15% from organic search - an eleven-fold advantage. HubSpot's 36% lift figure is narrower in scope - it covers purchase completion in a B2B CRM context specifically - but it points in the same direction.

What answer engine optimization means in practice

HubSpot uses the term answer engine optimization, abbreviated to AEO, to describe the practice of ensuring a brand appears prominently in AI-generated responses. The company formally launched an AEO product on April 14, 2026, priced at $50 per month as part of its Spring 2026 Spotlight. The June 12 data release is positioned as evidence of that tool's impact, not just of the trend in general.

According to HubSpot, customers actively using the AEO product are generating 20% more traffic from AI visits than comparable customers who are not using it. The differential in marketing-qualified leads is larger: 170% more MQLs. The differential in closed deals is also substantial: 82% more deals. HubSpot attributes all three figures to its proprietary customer data. None of the figures has been independently verified by a third party.

The AEO product itself tracks how often a brand appears in responses generated by ChatGPT, Gemini, and Perplexity. HubSpot subsequently launched an AEO Sensor feature, which provides ongoing monitoring of brand visibility across these platforms. The infrastructure behind the product traces to HubSpot's acquisition of XFunnel on October 31, 2025. XFunnel had, at the time of acquisition, analyzed 1,500 companies, collected over 5 million responses, and examined more than 25 million citations.

AEO, as HubSpot describes it, differs from traditional SEO in its inputs. Answer engines - ChatGPT, Gemini, Perplexity - decide whether to recommend a brand based on social media presence, third-party reviews, external websites, and owned content. Static keyword optimization does not directly influence those decisions. The discipline therefore requires tracking where a brand appears in AI responses, understanding which signals correlate with visibility, and adjusting the brand's external digital footprint accordingly.

The mechanics of AI search as a discovery layer

The June 12 publication describes AI search as functioning as a new discovery layer - one that decreases research time, changes how buyers validate options, and rewards brands that are optimized for conversational queries rather than static keywords. That framing points to a structural change in the buyer journey rather than a seasonal or cyclical one.

Traditional search engines return a ranked list of links. The buyer selects from those links, visits websites, reads content, and forms a view. The process involves multiple steps and multiple decisions. AI search compresses that process. A buyer asking ChatGPT which CRM to consider receives a synthesized answer that already incorporates comparison, prioritization, and sometimes recommendation. Fewer steps. Less time. Potentially fewer vendor touchpoints.

For marketing teams, that compression changes the question. The question is no longer only "how do we rank for relevant keywords?" It becomes "do we appear, at all, in the answers AI platforms generate for relevant queries?" A brand absent from AI answers may not make the shortlist before a buyer has even opened a browser tab with the vendor's website.

PPC Land has documented the commercial intent dimension of this shift: data from October 2025 showed that 21.6% of ChatGPT interactions demonstrate some degree of commercial intent, with 7.1% showing high intent with strong purchase signals. Business finance accounted for 34% of high-intent conversations. Product comparison represented 28%. Those figures predate the mass adoption of AI search in formal purchasing processes; the HubSpot survey data suggests the commercial use of AI search among B2B buyers has grown substantially since then.

The measurement problem

One complexity that HubSpot's data does not fully address is attribution. Tracking the influence of AI search on a purchase decision requires knowing that a buyer used AI search during their evaluation - something that is difficult to capture in standard web analytics. A buyer who asks ChatGPT about CRM options, forms a view, and then visits a vendor's website directly will appear in analytics as direct traffic, not as an AI referral.

This attribution gap has been a recurring challenge. Google Analytics added a dedicated AI assistant channel for traffic from ChatGPT, Gemini, Claude, and Perplexity, making it easier to identify direct AI referrals. But that channel only captures buyers who clicked a link from within an AI platform. Buyers who researched in AI tools and then navigated directly to a vendor remain largely invisible in standard reporting.

HubSpot's approach - asking buyers directly, in a survey, whether they used AI search during their evaluation - sidesteps the technical attribution problem. It captures self-reported behavior rather than observed behavior. The 42% figure may therefore be more accurate than what any analytics platform could derive from referral data alone, though self-reported data carries its own limitations around recall accuracy and social desirability bias.

Context: a fragmented and fast-moving market

The AI platforms most commonly associated with this discovery shift are not static. Data from December 2025 showed ChatGPT holding 66% of the U.S. AI chatbot market by traffic share, with Google Gemini climbing to 22% of global AI website traffic. By January 2026, that picture was already shifting, with Gemini surging and ChatGPT's lead narrowing.

The fragmentation matters for marketers. A brand that optimizes for ChatGPT visibility but not Gemini is not fully covered. HubSpot's AEO product tracks all three major platforms - ChatGPT, Gemini, and Perplexity - which reflects the multi-platform nature of the problem.

Meanwhile, the broader organic search environment has been deteriorating independently of AI chatbot usage. Google AI Overviews have reduced organic click-through rates on informational queries by 61% since mid-2024, according to research from Seer Interactive published in November 2025. In Germany, SISTRIX data showed 265 million lost organic clicks per month attributable to AI Overviews at position one. These are separate mechanisms from AI chatbot search - one involves AI-generated content within Google's own results, the other involves buyers leaving Google entirely - but both reduce the volume and quality of organic traffic that reaches brand websites.

The HubSpot survey does not directly address Google AI Overviews. Its AI search category refers specifically to standalone conversational AI tools. But for marketing practitioners, the two trends compound: traditional organic traffic is shrinking from two directions simultaneously.

What the numbers mean for marketing strategy

The HubSpot publication does not prescribe marketing strategy, but the data it presents has clear implications for how marketing teams allocate attention and budget. A 36% lift in purchase probability associated with AI search usage means that buyers who find a brand through AI platforms are higher-quality leads, on average, than those who do not use AI search. A 20% traffic lift, 170% MQL lift, and 82% deal lift among AEO-active HubSpot customers are figures that, if they hold up in independent analysis, would justify significant reallocation.

The discipline itself remains new. According to HubSpot, AEO is a brand new discipline with no established playbook. Answer engines decide whether to recommend a brand based on social media presence, reviews, third-party websites, and owned content. That input set overlaps with but does not replicate the signals that drive traditional SEO. A brand with strong Google rankings may have weak AI search visibility, and vice versa.

HubSpot and Reddit research published in January 2026 documented why B2B buyers are increasingly bypassing company websites, relying instead on community content and third-party validation. AI search amplifies that dynamic: it surfaces the community signals and review data that already influence buyers, then presents them in synthesized form before a buyer has ever visited a vendor's website.

The June 12 data release from HubSpot is self-interested - the company is making the case for a product it sells. But the underlying figures - a 27% organic traffic decline across nearly 300,000 customers, a 42% AI search adoption rate among CRM buyers, a 36% lift in purchase probability - are specific and large enough to take seriously as signals of a structural shift in B2B buyer behavior, whatever one concludes about the AEO product category itself.

Timeline

Summary

Who: HubSpot, a marketing and CRM platform serving nearly 300,000 businesses, published the data. The survey covered more than 3,000 CRM purchase decision-makers worldwide.

What: HubSpot's January 2026 survey found that AI search - using platforms such as ChatGPT, Claude, and Gemini during a purchasing evaluation - is the single strongest predictor of CRM purchase intent, ahead of product demos, review sites, and sales calls. Buyers who used AI search were 36% more likely to purchase. HubSpot also reported that customers using its AEO product generate 20% more AI traffic, 170% more MQLs, and 82% more closed deals than comparable customers not using it. Organic traffic across HubSpot's customer base has declined 27% year over year.

When: The data was published on June 12, 2026. The underlying survey was conducted in January 2026, covering more than 3,000 respondents.

Where: The publication appeared on HubSpot's company news blog. The survey covered CRM purchase decision-makers worldwide, across multiple geographies.

Why: The findings matter because they quantify the purchase-intent impact of AI search at a scale and specificity not previously published in the B2B context. A 36% lift in purchase probability is large enough to influence marketing budget allocation, measurement frameworks, and content strategy. The data also reinforces the case for answer engine optimization as a distinct discipline, separate from traditional SEO, at a moment when organic traffic is declining across both traditional search and AI-intercepted queries.