How to measure the dark funnel

You can't track the dark funnel touchpoint by touchpoint, but you can measure its size. The method: read the five observable traces that hidden buying activity leaves in systems you already own, score them against benchmarks, and express the result as a dollar gap. Here's the full approach, usable with or without us.

What is the dark funnel?

The dark funnel is every buying touchpoint your attribution can't see: AI chatbot research, peer recommendations in Slack and WhatsApp, community threads, podcast listens, anonymous site visits, review-site lurking. It's most of the journey.6sense found that B2B buyers have largely made up their minds before they first contact a vendor, with 84% of deals going to the vendor they already favored.

Dark funnel vs. dark social

Dark social is one part of the dark funnel. It means sharing you can't track: links pasted into Slack, WhatsApp, text messages and email. The dark funnel is bigger. It also covers AI chatbot research, podcasts, review sites, events and anonymous visits to your own site. Fixing dark social tracking still leaves most of the dark funnel unseen.

Why attribution software can't fix this

Attribution tracks identified people on owned properties. The dark funnel is anonymous people on properties you don't own. No pixel fixes that. Teams that respond by buying more attribution tooling get more precise reporting on the shrinking slice of the journey they could already see.

How do you measure the dark funnel?

You measure the dark funnel by its traces, not its touchpoints. Read five signals in systems you already own: anonymous ICP engagement, AI discovery, branded search gap, repeat evaluation and peer influence. Score each against a benchmark, add them into a 0 to 25 score, and turn the score into a dollar gap.

The five measurable signals

Hidden demand is invisible directly but loud indirectly. Five traces, and where to read each:

1. Anonymous ICP engagement. In your account-ID platform (6sense, Demandbase, or similar), what share of right-fit account visits never becomes a known contact? High anonymous share means heavy off-funnel evaluation. No platform? Use the ratio of ICP-fit traffic (by firmographic enrichment) to identified ICP leads as a rougher read.

2. AI/LLM discovery. Ask ChatGPT, Perplexity, Copilot, and Gemini the eight questions buyers actually ask in your category. Count how many answers name you versus a competitor. Fully reproducible; the exact method is here.

3. Branded search gap. Compare your branded search impression trend (Search Console) to your captured-demand trend (CRM). Brand interest rising faster than conversions means demand is forming that your funnel isn't collecting.

4. Repeat evaluation. In analytics, how many sessions does a converting visitor average before their first conversion? Long anonymous evaluation stretches mean the decision is happening before you get a name. Be honest about precision here: without account-ID tooling you can't cleanly filter to ICP traffic, so treat this one as directional.

5. Peer influence. How often does your name come up in communities, review sites, and "how did you hear about us" answers? Triangulate from public sources plus a self-reported attribution field on every form. If you don't have that field, add it this week; it's the cheapest dark-funnel instrument that exists.

Turning the signals into a number a CFO accepts

Score each signal against a benchmark, sum to a 0-to-25 Demand Visibility Score, and map the score to a benchmark range for influence investment. Then one subtraction: what the benchmark says your influence budget should be, minus what it is. That gap, in dollars, is the measurement. Not a projection, not modeled revenue. Your budget times a cited benchmark, minus your spend. Pipeline implications can be layered on as a range (2x to 4x), but the subtraction is the number nobody can argue with. How that gap actually becomes pipeline is here: how invisible demand becomes new pipeline.

Doing it yourself vs. running the diagnostic

Everything above is doable in-house; that's the point of publishing it. The Invisible Demand Diagnostic is the packaged version: 45 minutes, five inputs, scored against published benchmarks, with the gap calculation and a funded roadmap ranked by your weakest signals. The first assessment is free. Either way, measure before you spend. The most common failure in dark-funnel work is funding the fashionable fix instead of the dark channel.