Revenue leakage.
Two leaks, one bucket: dark-funnel demand you never see, and pipeline you saw but stalled. Different problems, both unmeasured.
Revenue leaks in two distinct places. Demand that's there but invisible to you (dark social, AI shortlists, peer referrals). And pipeline you sourced but never closed (stalled deals, ghosted champions, late-stage no-decisions). Most orgs measure neither — and conflate them when they try.
Four symptoms we see, almost every time.
- 01
'Where did this deal come from?' has no good answer
- 02
Pipeline coverage looks fine; conversion is the problem
- 03
Stalled deals are reported as 'in progress' for quarters
- 04
No closed-lost loop back into marketing
Of pipeline influenced by un-attributable dark-funnel activity (6sense)
Of forecasted deals end in no-decision (Gartner)
Three moves — in this order, every time.
- 01
Self-reported attribution
'How did you hear about us?' on every inbound, reconciled against MTA.
- 02
Stall diagnostics
Stage-by-stage exit analysis to find where deals actually die.
- 03
Closed-lost loop
Structured win/loss feeding messaging, ICP and product — quarterly, not anecdotal.
What this gap actually costs.
Revenue leaks in two unrelated places, and most reporting blends them. The first is demand that exists but is invisible to you: peer conversations, private communities, analyst chatter and AI-generated shortlists that no attribution model sees. The second is pipeline you created and then lost to inertia rather than to a competitor.
Conflating them leads to the wrong fix. Invisible demand is a measurement and presence problem; stalled pipeline is a qualification and process problem. Teams that treat both as a lead-volume problem spend more at the top of the funnel and watch conversion fall again.
No-decision is the largest single competitor in enterprise technology. Deals do not get lost so much as postponed — the business case fails a budget review, the sponsor moves, the risk of changing exceeds the risk of staying. None of that shows up in a competitive win-loss field.
The version that works.
- 01
Self-reported attribution on every inbound, reconciled against platform data rather than replaced by it.
- 02
Stage-exit analysis that shows where deals actually die, with no-decision tracked as its own outcome and its own reasons.
- 03
A structured closed-lost loop with named owners, running quarterly, feeding messaging, ICP and product rather than sitting in a deck.
- 04
Forecast hygiene: stalled deals reclassified honestly instead of carried forward as coverage.
Four ways good teams get this wrong.
Attribution as truth
Treating a multi-touch model as a complete account of influence hides the majority of enterprise demand creation, which happens where no pixel fires.
No-decision recorded as lost to competitor
Mislabelling the outcome removes the single most useful diagnostic signal you have.
Coverage theatre
Pipeline coverage ratios look healthy while stalled deals inflate the denominator. The ratio reassures; the conversion rate does not.
Anecdotal win/loss
Loss reasons collected by the rep who lost the deal, in a free-text field, at the end of a quarter, produce a pattern nobody can act on.
- 01
Add self-reported attribution and compare it against the existing model for one quarter — the delta is your dark-funnel estimate.
- 02
Run a stage-exit analysis on 12 months of opportunities, separating competitive losses from no-decisions.
- 03
Stand up the closed-lost loop with named owners and a quarterly forum where the findings change something specific.
- 01Dark-funnel measurement model
- 02Pipeline stall + no-decision report
- 03Closed-lost loop with named owners
You see the demand you were missing, plug the stalls you were ignoring, and stop confusing the two.
Book the diagnosticFrequently asked.
- What exactly is the dark funnel?
- Demand creation and research that happens on surfaces you cannot measure: peer recommendations, private communities, podcasts, analyst conversations and AI-generated shortlists. It is real influence with no trackable touchpoint.
- How do you measure something invisible?
- Indirectly and honestly. Self-reported attribution at point of enquiry, combined with branded-demand and citation trends, gives a directional read that is far more useful than pretending the channel does not exist.
- Is a stalled deal the same as a lost deal?
- Operationally, yes — but the cause is different. Stalls usually indicate a weak business case or a missing stakeholder, which is a fixable process problem rather than a competitive one.
- Who should own the closed-lost loop?
- Marketing usually runs it, with sales participation and a standing slot in the leadership cadence. Without a named owner it reverts to anecdote within a quarter.