How to Know If Your Demand Gen Is Working (When Your CRM Lies to You)
Where did your last ten customers first hear about you?
Check the CRM and a lot of them will say Direct. That's the problem with judging demand gen by the CRM. It only knows what it's been told, and in most B2B companies it's been told a lot less than what happened.
The LinkedIn post someone forwarded to a colleague. The podcast that made a VP search your name. The recommendation in a Slack group you'll never see. None of it lives in Salesforce. So your attribution model, however well built, is an incomplete picture of why pipeline grows.
So measure with your eyes open.
The metrics that tell the truth
Pipeline, not leads. If MQL volume is your main demand gen number, you're one step away from the one that matters. MQLs are an input, and pipeline is what you're measuring. A month with 200 MQLs and $0 in pipeline is not a good month, whatever the chart says.
Track the dollar value of opportunities where marketing was the first touch or a documented influence. That's what tells you whether demand gen is working.
Stage conversion rates, trended. One snapshot tells you where you are. A trend tells you if it's getting better or worse, and often points at the change that did it.
MQL to SQL, SQL to opportunity, opportunity to closed won. Check them monthly. When one drops, something changed upstream: lead quality, the sales handoff, or a channel sending the wrong people.
Time to pipeline. Days from first marketing touch to an open opportunity. If it's growing, something is adding friction, and if it's shrinking, something in nurture or conversion is working. It's an early signal. It tells you about pipeline before it shows up in the pipeline number.
Sourced and influenced, kept apart. Sourced means marketing was the first touch. Influenced means marketing touched the account before the opportunity was created. Sourced tells you about acquisition. Influenced tells you about marketing's role across the whole buying cycle, which is usually much bigger. Report both, in separate columns.
Once a quarter, ask sales what prospects mention on first calls, and write it down.
The dark funnel
Even perfect attribution won't catch everything. A lot of B2B buying happens where you can't track it: private communities, DMs, review sites, word of mouth, conversations inside the prospect's company that you're not in.
That's the dark funnel: buying intent that leaves no trace in your systems. It drives pipeline whether or not the CRM can see it.
You can still add what people tell you to what the systems record.
Put a "How did you hear about us?" field on your high-intent forms, and make it an open text field. The answers will surprise you. "My former colleague mentioned you." "A LinkedIn post from someone on your team." "I've been reading your blog for six months." None of those show up in your attribution model. Reading them by hand doesn't scale, but it tells you where attention is coming from.
When the numbers don't agree
Say your attribution says paid LinkedIn makes almost no pipeline, but the "how did you hear" answers keep saying LinkedIn. Look closer before you cut it.
People don't separate a paid ad from an organic post or a colleague's share. The answers name the platform and rarely the campaign. It might be a tracking gap. It might be organic doing the work. Check the paid tracking first, then decide.
Channels get cut this way. The CRM shows nothing, the channel goes, and three months later pipeline dips and nobody can say why.
So cross-check the attribution against whatever people are telling you. Then check the plumbing: a subdomain tracking gap, missing UTMs, a conversion event that isn't firing. The gap may have a technical explanation.
Before you cut a channel, read the last 50 "how did you hear about us" answers.
What good measurement needs
It takes three things working together.
Clean data. UTMs on everything, hidden fields on every form, cross-domain tracking set up, HubSpot and Salesforce syncing lead source and lifecycle stage. Without this, every metric is a guess.
The right metrics. Pipeline, stage conversion, time to pipeline, influenced revenue. Sessions, raw MQL volume and open rates stay off the list.
Humility. Self-reported answers, regular sanity checks against what sales is hearing, and no channel decisions based only on attribution you already know is incomplete.
Measurement is never finished, and the dark funnel stays dark. Keep reading the "how did you hear" answers every month, even after the dashboards look clean.
Not sure your tracking holds up?
Run the free Attribution Audit
About five minutes, 21 questions. You get a score and the likely places your pipeline data breaks, from UTMs and forms to CRM sync and reporting.
Run the free auditNotes from the Field.
Every other week, one broken thing and how I’d fix it.
One email every other week · unsubscribe anytime · Privacy policy