Gartner’s 2026 research puts 70–80% of the B2B purchase journey before any vendor contact. 61% of buyers actively prefer a rep-free evaluation. 92% walk into vendor conversations with a preferred choice already in mind — a preference formed in private Slack groups, peer DMs, podcasts, analyst conversations, and ChatGPT queries that will never appear in any dashboard you can build. The default response is to invest harder in the channels you can track, because at least those produce a number. The problem is that the channels you can track are not the channels that matter to the decision.
TL;DR
- The measurable funnel and the actual buyer journey have diverged. Most B2B teams are still trying to optimize the visible 25%.
- Self-reported attribution is the most honest signal a marketing team can capture, and the most consistently dismissed by CFOs who call it “fluffy.” Making it credible is an operational discipline, not a methodology problem.
- Branded search velocity, direct traffic, podcast and content reach, and share-of-model in LLM answers are the leading indicators that the dark funnel is working. None of them attribute neatly to a campaign — and that’s the point.
- The diagnostic that a team is faking dark-funnel insight rather than working it: every dark-funnel data point gets explained back into a channel attribution. Real dark-funnel measurement leaves results sitting outside the attribution model on purpose.
What the Dark Funnel Actually Is
The dark funnel is not a clever way of describing channels you don’t track. It’s the structural fact that most B2B decision-making now happens in environments that are technically untrackable from the vendor’s side: peer Slack channels, LinkedIn DMs, analyst calls, podcast listens, ChatGPT and Perplexity conversations, internal champion-built spreadsheets, vendor RFI shortlists assembled before a single rep is involved.
Forrester and 6sense data converge on the same picture. By the time a buyer fills out a form on your site, they’ve usually read 5–10 pieces of content, listened to or watched 2–4 longer-format conversations about the space, asked an AI model to summarize the category, and discussed the shortlist with peers. The form fill is the moment the funnel becomes visible. It’s nowhere near the moment the decision started forming.
Why Self-Reported Attribution Is the Most Honest Signal
The cleanest data point on dark-funnel performance is the question on the demo-request form: “How did you hear about us?” Marketers undervalue it because the data is messy — buyers misremember, conflate channels, name the most recent touchpoint, leave fields blank. It is still the only direct measurement of what the buyer believes drove them in.
Aggregated over enough volume, self-reported attribution becomes credible. Companies that have run this for 12+ months see consistent patterns: 30–50% name a podcast, a specific LinkedIn creator, a peer recommendation, or a piece of long-form content that nothing else in the attribution stack ever captured. None of those channels would show meaningful credit in last-touch or multi-touch attribution because the touchpoint did not produce a click that resolved to a session that converted.
The way to make self-reported attribution credible to a CFO who calls it fluffy: aggregate it weekly, normalize it across deal segments, compare the named channels to the actual time-lagged change in branded search and direct traffic, and publish a single number monthly — the share of self-reported demand that came from channels the attribution dashboard doesn’t credit. When that number exceeds 30%, the conversation stops being about whether dark funnel exists and starts being about whether the team is funding it.
The Indicators That the Dark Funnel Is Working
No single metric captures dark-funnel performance because the dark funnel isn’t a channel. It’s the cumulative effect of being present in the conversations buyers have when no vendor is in the room. The leading indicators that show up consistently across companies that build this well:
Branded search velocity. If branded search volume is rising faster than total category search volume, your share of demand inside the dark funnel is growing. If it’s flat or falling while category volume rises, you’re being talked about less than you used to be.
Direct traffic and unattributed sessions. Direct traffic to specific high-intent pages (pricing, comparison, integration pages) is a near-pure dark-funnel signal. Someone heard about you, formed enough intent to type the URL, and arrived without any tracked source. Total direct traffic is noisy. Direct to a comparison page is not.
Share of model. A growing share of B2B research now starts with ChatGPT or Perplexity. You can measure your share of model by running a stable set of category-relevant prompts on a regular cadence and tracking which vendors get named, in which order, in which context. This isn’t a vanity metric — it’s the leading indicator of which vendors will land on the procurement shortlist before any rep is engaged.
Sales-conversation discovery. The rep’s first call is the only point where the buyer narrates their own journey. Capturing “what brought you in” and “who else are you looking at” as structured fields in CRM, then trending those over time, is the closest thing to a primary-source measurement of dark-funnel reach.
Where Most Teams Fake It
Faking dark-funnel measurement looks like this: every quarter, the team produces a slide showing “branded search up 12%, direct traffic up 8%, podcast downloads up 27%, AI Overviews mentions up 4x.” The slide is real. The diagnostic is what happens next. Does the budget move? Does the content roadmap change? Does the executive thought-leadership investment increase? Does anyone get held accountable for the share-of-model number?
If the answer is no — if dark-funnel data is harvested for a slide and the operating plan continues unchanged — the team is performing dark-funnel measurement, not running it. The disqualifier is that every dark-funnel data point gets folded back into a channel attribution. “Branded search is up because of our paid LinkedIn campaign.” “Direct traffic is up because of the conference.” Real dark-funnel insight leaves room for the answer to be: we don’t know which specific investment caused this, but we can see the cumulative system is working — or it isn’t.
Where Marketing’s Accountability Has to Move
The shift this forces on a marketing function is uncomfortable. The accountability moves from sourced pipeline (which the dark funnel structurally undercounts) to a smaller number of leading indicators that explicitly cannot be tied to a specific campaign. Branded search growth. Share of model on a defined prompt set. Direct-to-high-intent-page traffic. A monthly self-reported attribution number. Pipeline win-rate differentials between deals where the rep heard “I read your stuff for a year” versus “I clicked your ad last week.”
The CMO who wants the dark funnel to count needs to negotiate the scorecard with the CFO directly. The compromise that holds: a small set of dark-funnel leading indicators, agreed in advance, reviewed monthly, given equal weight to sourced pipeline in the operating review. The CFO won’t sign off on dark-funnel measurement in the abstract. They will sign off on a specific scorecard that produces specific numbers on a known cadence.
The Bottom Line
The visible funnel is shrinking. The decisions that determine whether you land on a shortlist are forming somewhere you cannot directly observe. Trying to optimize harder against the visible 25% is the most expensive habit in B2B marketing right now, because it starves the investments that are doing the actual work. The teams that get ahead of this are the ones that pick three or four leading indicators they cannot attribute, defend them as the operating metrics, and let the attribution dashboard be useful for what it is — a tactical optimization tool — rather than the source of truth it was never built to be.
Additional Resources
From the Zaitz Marketing Knowledge Library:
- How to Measure the AI Dark Funnel in B2B — The specific case of LLM-driven research before vendor contact
- B2B Lead Scoring for Invisible Buyers — How scoring breaks when the journey is unobserved
- Why Your Attribution Model Is Lying to You — The structural reasons attribution misses the dark funnel
Want a second read on your measurement setup?
Start with a Growth Architecture Review. We will map your channel mix, audit your attribution, and show you where the real leverage is.
Book a Conversation →