AI Summary
The B2B buyer journey in 2026 is mostly invisible: buyers use AI tools to research vendors, build comparison matrices, and reach consensus before ever contacting sales, which means your GEO and content strategy now determines deals that your CRM will never show you.
How AI tools changed the self-serve research phase
The traditional B2B buying process involved Google searches, analyst reports, and direct visits to vendor websites. In 2026, that process is increasingly mediated by AI tools: buyers ask ChatGPT, Claude, or Perplexity to compare the top five project management platforms for remote teams, get a synthesized answer, and never click through to a single vendor site.
Forrester’s B2B Buyer Adoption of Generative AI report (November 2024) found that 89% of B2B buyers have adopted generative AI and name it one of the top sources of self-guided information in every phase of their buying process. At 89% adoption, this is a mainstream shift. When a buyer asks an AI for a vendor comparison, the AI synthesizes information from multiple sources and presents a summary. The buyer may never visit your website, never appear in your analytics, and never engage with your sales team, yet they still form a purchasing opinion based on what the AI told them.
The structural implication: your brand must appear correctly in AI-generated answers across multiple query types, not just a single category search. We cover the full optimisation playbook in what is GEO and how it works. The buyer journey post you are reading now focuses on the GTM strategy layer: how to influence buying decisions when most of the journey is invisible.
The dark funnel: 70% of buying happens before sales contact
The dark funnel refers to buyer touchpoints that occur outside of trackable attribution: Slack community discussions, podcast listening, word-of-mouth referrals, Reddit threads, and in 2026, AI-mediated research via ChatGPT, Perplexity, and Claude. 6sense research on B2B buyer behavior found that serious buyers do not speak with vendors until they are 70% through their buying journeys. By the time a prospect fills out a demo form or replies to a sales email, they have already formed opinions, evaluated competitors, and built internal consensus, all in channels you cannot attribute.
AI tools have accelerated dark funnel expansion in a specific way. Before AI, dark funnel research was human-mediated: you could influence it through community engagement and thought leadership because the conversations happened in public forums. In 2026, a significant portion of dark funnel research happens inside a private AI session. When a buyer asks ChatGPT about your product, you have no visibility into the query or the answer unless you proactively test AI responses to your category’s prompts.
This AI-mediated dark funnel is also multi-stakeholder. Forrester’s State of Business Buying 2024 report found that on average 13 people within an organisation are involved in the buying decision, and that 86% of B2B purchases stall during the buying process as those groups struggle to reach consensus. Each stakeholder conducts independent AI-assisted research, and almost 95% of buyers anticipate using generative AI to support their purchase process in the next 12 months. Your brand must appear correctly across multiple concurrent AI queries from different roles, not just the champion’s single search.
Dark funnel signals you can track in 2026
The dark funnel is invisible to traditional attribution, but there are downstream signals that indicate activity inside it. These are consistent with what we see in our GA4 AI search referral attribution work, where AI referrer fragments are catchable but account for only a fraction of total AI-driven interest.
- Branded search lift. A sudden increase in branded searches for your product name often indicates that buyers encountered your brand in a dark funnel channel, including an AI-generated answer, and are now conducting follow-up research. Monitoring branded search volume trends is the most reliable dark funnel proxy available in standard analytics.
- Direct URL visits. When a user types your URL directly or uses a bookmark, they were likely referred by a source that did not provide a clickable link: a podcast mention, a Slack message, or an AI-generated recommendation. Spikes in direct traffic correlated with specific campaigns or content launches are a strong dark funnel signal.
- AI referrer fragments. Some AI platforms pass referrer data that appears in server logs or GA4 as partial fragments. Our May 2026 study of 153,425 citations found these fragments are measurable but undercount real AI-driven traffic significantly, since most AI sessions generate no referrer at all.
- Demo requests with no prior attribution. Inbound requests from companies that have no prior touch in your CRM are often dark funnel arrivals. Ask the first qualifying question: ‘How did you first hear about us?’ The answer is frequently ChatGPT or Perplexity.
- AI crawler activity on commercial pages. Monitor your server logs for OAI-SearchBot, PerplexityBot, and Claude-SearchBot hits on pricing and feature pages. High-frequency crawls of those pages precede citation activity. See our AI crawler control guide for the full log analysis setup.
Attribution in the dark funnel era: what you can and cannot measure
Traditional attribution models assume you can track the buyer journey from awareness to conversion. In 2026, that assumption fails for the majority of the journey. The honest framing: you cannot attribute what you cannot see. Measure share-of-voice proxies instead of trying to force attribution onto invisible touchpoints.
| Signal | What it measures | Reliability |
|---|---|---|
| Branded search volume | Dark funnel brand awareness proxy | High: consistent and directly measurable |
| Direct traffic trends | Unlinkable referrals (AI, podcast, Slack) | Medium: noisy, correlated not causal |
| AI referrer fragments in GA4 | Fraction of AI-driven clicks (not sessions) | Low: undercounts by design; most sessions have no referrer |
| AI crawler log hits on commercial pages | Indexing activity that precedes citations | Medium: leading indicator, not conversion data |
| Share of voice on AI recommendation prompts | Brand presence across AI answer engines | High: most direct proxy for dark funnel AI exposure |
| First-touch qualitative (‘how did you hear about us?’) | Actual awareness channel for new buyers | High: small sample but actionable |
Share of voice on AI recommendation prompts deserves special attention because it is the closest measurable proxy to what happens inside a dark funnel AI session. Track share of voice in AI search by running scripted prompts weekly across the major AI engines in your category and logging which brands get named. Our GEO/AEO Tracker automates this across ChatGPT, Perplexity, Claude, and Google AI Overviews. Correlate share-of-voice trends with branded search volume and pipeline creation for the most complete dark funnel picture available.
Influencing invisible buyers: three levers
You cannot track the dark funnel, but you can influence it. The goal is to ensure that when buyers ask AI tools for vendor recommendations, your brand appears in the answer, and that when they encounter your brand in a dark channel, they arrive on a site that converts.
Lever 1: AI visibility as demand creation. Optimize your content for AI citation so that when buyers ask ChatGPT or Perplexity for vendor comparisons, your brand appears in the answer. This is the core of generative engine optimisation (GEO). Our May 2026 study of 153,425 citations found that 76.95% of cited URLs are not in the organic top-10, meaning GEO reach extends well beyond traditional SEO. The cited sentence length data matters too: that same study found the mean cited sentence was 9.27 words, with 45.2% of all cited sentences in the 6-10 word range. Write atomic factual statements about your product capabilities to match that citation pattern. Full methodology is in our 153,425-citation study.
Lever 2: Content that travels in dark channels. Original research, comprehensive comparison guides, and contrarian analysis get shared in Slack, forwarded to colleagues, and referenced in private conversations. Those shares do not generate attribution but they build awareness in the dark funnel. Our decoded 42,971 AI citations study is a direct example: it gets cited by AI engines because it contains primary data, and it gets shared in practitioner communities because it is genuinely useful. See the primary research as AI authority signal post for the content type analysis.
Lever 3: Comparison content that AI can parse. When buyers ask AI to compare you to a competitor, you should own a page that answers that query directly. AI systems prioritise structured comparison content. A clear feature table with consistent terminology, published under a ‘ProductA vs ProductB’ URL, gets extracted and cited. Buried prose comparisons do not. Our comparison page template guide covers the exact structure AI search engines cite most.
Agentic buying: the next layer of invisible purchase activity
The dark funnel is about to get darker. AI agents like ChatGPT agent, Claude-User, and Perplexity Comet are already browsing vendor websites, extracting pricing and feature data, and building comparison matrices on behalf of buyers. This represents a new category of invisible buying activity: invisible research, invisible evaluation, and potential transaction initiation.
The dark funnel implication is straightforward: the same buyer who used to research your product privately in ChatGPT is now sending an AI agent to browse your site directly. The agent does not fire your tracking scripts, does not appear in GA4, but does parse your schema, evaluate your pricing transparency, and include or exclude you from the shortlist. Full optimisation patterns are in our agentic search optimisation guide.
GTM strategy for the AI-mediated buyer journey
The shift to AI-mediated buying forces a reorientation from demand generation to demand capture. Buyers generate their own demand by asking AI tools for recommendations, and your GTM job is to be in the answer when they ask. Our GTM strategy service is built on this premise: B2B companies that invest in AI visibility capture disproportionate demand because they are recommended by AI systems across the dark funnel, while competitors who ignore this channel lose share.
- Allow AI retrieval crawlers (OAI-SearchBot, PerplexityBot, Claude-SearchBot) in robots.txt on all commercial pages
- Publish transparent pricing so AI can cite your cost structure in buyer research queries
- Publish comparison pages that answer ‘ProductA vs ProductB’ queries with structured feature tables
- Monitor share of voice vs search volume as your leading dark funnel KPI
- Track branded search trends weekly as the most reliable dark funnel signal
- Ask new customers ‘how did you first hear about us?’ and log the answer: direct qualitative feedback is the most actionable dark funnel data you have
The metric to watch is AI share of voice on recommendation prompts in your category. Run scripted prompts weekly across the major agents and log which brands get named. Our GEO/AEO Tracker automates this across ChatGPT, Perplexity, Claude, and Google AI Overviews.