The 'how did you hear about us' survey, done right: a playbook

The 'how did you hear about us' survey, done right: a playbook

The question: self-reported attribution is famously biased — so how do you run a source survey that's actually worth trusting?

Self-reported attribution (asking buyers directly) is the main way to catch dark social and word-of-mouth that analytics miss. But naive surveys produce garbage, skewing toward whatever the buyer saw most recently.

The playbook to de-bias it:
— Ask at the moment of highest intent — on the demo form, not a delayed email.
— Use an open text field, not a dropdown; dropdowns plant answers.
— Keep it to one question; every added field cuts completion.
— Reconcile self-report against analytics, and flag systematic gaps (if 40% say 'a podcast' your tracking misses, that's your dark channel).
— Tag the recency bias: recent touches get over-credited, first touches forgotten.
— Review quarterly for shifts in the open-text themes.

Limitations: humans misremember and compress multi-touch journeys into one source; small B2B deal volume makes percentages jumpy. Treat results as a triangulation input, not ground truth.

What it means for B2B: the survey's value isn't precision — it's catching channels your analytics structurally can't see.

Bottom line: use self-report to find your blind spots, not to settle credit disputes.