Book a call

Daniel & Friends / Blog / GTM & Sales

Dark funnel: the decisive part of the buying journey attribution cannot see

The dark funnel is the part of the buying journey that happens beyond the reach of analytics tools: referrals between colleagues, private communities, podcasts, newsletters and dark social, meaning links shared in chats and emails with no referrer. That is where buyers form their view. What reaches your CRM is a decision already made.

Daniel Votruba · 6 July 2026 · 4 min read

Why last click lies systematically

Attribution software measures clicks and cookies. The buying decision, however, forms elsewhere. Gartner research finds a B2B buying group spends just 17% of the time of the whole buying process with supplier sales reps; the rest goes on internal discussion and independent research. In its Buyer Experience Report, 6sense shows buyers contact suppliers only once they are roughly 70% decided, and in the overwhelming majority of cases the vendor they approached first wins. Add the 95:5 rule (only a sliver of the market is buying right now) and you get an uncomfortable truth: demand forms during a period when nothing is measuring it.

Last click, meanwhile, reliably captures the final step, typically branded search or a direct visit. The resulting report then suggests that brand Google Ads works and nothing else does. But that is not measurement, it is an optical illusion: the channel that harvested the demand claims credit from the channels that sowed it. A company that pours budget into harvesting and cuts the sowing on the strength of such a report is systematically switching off its own future growth.

Self-reported attribution: just ask people

The cheapest correction to that illusion costs one form field: "How did you hear about us?" Required, open, with no pre-filled options to steer the answer. While the software reports organic search, a person writes that a former colleague recommended you, or that they heard you on a podcast. Methodologically this is soft data, because memory is fickle and people recall the last touch they registered. Even so, systematically comparing self-reported answers with software attribution is one of the most instructive things a marketing team can do. The gap between the two pictures is roughly a map of your dark funnel.

And the dark is getting darker. Search engines answer directly on the results page, AI assistants assemble recommendations from content that never produces a click, and social networks punish outbound links. An ever larger share of the influence on buying decisions therefore happens in environments where a pixel has nothing to measure. That is no reason to give up, but it is time to stop pretending the dashboard shows the whole of reality.

How to feed the dark funnel

  • Content worth sharing: proprietary data, opinions with an edge, genuinely usable guides. Nobody forwards an average article into a company Slack.
  • Presence in communities: wherever your buyers actually talk, meaning industry Slacks, LinkedIn, conferences. Contribute, do not sell.
  • Podcasts and newsletters: formats with long exposure and high trust. An hour in a listener's ears beats a thousand banners.
  • Brand: mental availability is exactly what makes a buyer think of you when the buying window finally opens.
  • Customer experience: the strongest dark funnel channel is happy clients who recommend you when you are not in the room.

The common denominator: all of these are hard to report in a quarterly deck and show up beautifully in the annual accounts. The dark funnel rewards companies able to defend an investment without immediate attribution proof. Which in the end is a question of leadership culture rather than marketing technique.

Measure approximately, but honestly

You will not measure the dark funnel to a tenth of a percent. Anyone promising that is selling a report, not the truth. You measure it by triangulation: self-reported attribution, the trend in branded search and direct traffic, community mentions, de-anonymisation of company visits to your site, and, on larger budgets, econometrics (marketing mix modeling or geo tests). Approximately right beats precisely wrong. The budget decisions that follow tend to be uncomfortable, because some of the money goes to channels with no immediate ROAS. Which is precisely why only a minority make them. And precisely why they work.

Want to know how big your dark funnel is and what to feed it? Here is how we work with intent and invisible demand. Or just write to us. The first opinion is free.

Key takeaways (TL;DR)

  • Buyers contact suppliers only when they are roughly 70% decided, and spend just 17% of the buying process with sales reps. Demand forms where nothing measures it.
  • Last click is an optical illusion: the channel that harvested demand takes credit from the channels that sowed it. Cut the sowing on that basis and you switch off future growth.
  • The cheapest correction is a required open field asking how they heard about you. The gap between people's answers and software attribution is roughly a map of your dark funnel.
  • The dark funnel is fed by shareable content, community presence, podcasts, brand and customer experience. It is measured by triangulation, where approximately right beats precisely wrong.

FAQ

What is the dark funnel?

The part of the buying journey analytics cannot see: colleague referrals, private communities, podcasts, newsletters, dark social and content consumed without a click. That is where buyers form their view; what reaches the CRM is a finished decision, which attribution then wrongly credits to the last click.

Is last-click attribution useless?

No, it is just systematically biased. It reliably measures the final step, which is why it overvalues channels that harvest demand and undervalues those that created it. Use it to diagnose the harvest, not to allocate budget.

How do you measure the dark funnel?

By triangulation: a required open question asking how they heard about you, the trend in branded search and direct traffic, mentions in communities and podcasts, and on larger budgets MMM or geo tests. No single method is accurate on its own; together they give a usable picture.

Sources: Gartner, B2B Buying Journey · 6sense, Buyer Experience Report · LinkedIn B2B Institute, The 95-5 Rule · our own practice across 750+ projects.