How Accurate Is Self-Reported Attribution for Lead Source?

How Accurate Are "How Did You Hear About Us?" Answers?

"How did you hear about us?" answers are accurate about the things buyers noticed, such as a person who recommended you or an event where they met you, and weak about the things they didn't, such as an ad they scrolled past three times. That makes them unreliable as a complete record and valuable as a second one.

The pattern has a basis in psychology, where a widely cited review of research on how people explain their own choices found that people often can't report accurately what influenced them; asked why, they tend to give the explanation that seems plausible, and their reports are accurate mainly when the real influence was noticeable and an obvious cause.¹ The review wasn't about marketing, but our reading of it matches what lead-source answers do: a colleague's recommendation is noticeable and obviously a reason to call, so buyers remember it, while a display ad or a retargeting impression rarely is, so buyers credit whatever they did last, which is often "Google".

So the answers err in a predictable direction, under-reporting background advertising and over-reporting the final step, usually a search for your name, while catching word of mouth better than any tracking tool can. The Channel Selection step shows where these answers sit in a source of truth you control.

Should You Ask "How Did You Hear About Us?" If You Already Track Lead Source?

You should ask even with good tracking, because tracking records the click that brought someone to your site, and a buyer who heard about you from a peer, a podcast or an association meeting will still arrive through a click, usually on a search for your name. The tracking is right about that click and silent about what caused it.

Take a commercial HVAC contractor with about forty million dollars in annual revenue, which sells maintenance agreements to property management firms through Google search, LinkedIn ads and a local building owners' association.

Without the question, the association would have looked like it produced nothing, and it's exactly the kind of channel that gets cut first when budgets tighten. How far the ad platforms' own claims ran past the agreements the contractor signed is on How Do You Measure Digital Marketing ROI?, and why no change of attribution model would have found the association is on Which Attribution Model Should You Use?.

How Should You Ask "How Did You Hear About Us?" to Get Honest Answers?

Ask it as an open question, make it required at intake, and have a person ask it again on the first call, then sort the answers into categories afterward. Where and how you ask decides most of how accurate the answers are.

1. Use open text on the web form, rather than a dropdown. A dropdown makes it easy to pick the first option or the most familiar one, and "Google" or "Search" is familiar to everyone because it's often how they found your number, whatever sent them looking.
2. Make the field required at intake, in the CRM record your team completes for every new lead, so no lead reaches a salesperson with the source blank.
3. Have a person ask on the first call, and follow up on vague answers: "Google" becomes "What were you searching for, and had anyone mentioned us before?" A conversation recovers the colleague, the event or the podcast that a form answer skips.
4. Ask early, at the first conversation rather than at signing, because memory of what started the search fades over a long sales cycle.
5. Keep the buyer's words and add a category, so "a colleague mentioned you at the association lunch" is stored as written and also coded as "association event" for reporting.
6. Keep the category list short and fixed, with referral, event or association, podcast or media, search, social, ad and other, so the monthly counts can be compared over time.

If you're a smaller business, steps 3 and 5 are most of the work, and a spreadsheet column with the answer in the buyer's words will do until a CRM is worth the cost.

How Do You Reconcile Self-Reported Answers With CRM Tracking?

Keep the tracked source and the self-reported source in two separate fields on every lead and opportunity, and never let one overwrite the other, because they answer different questions: how the buyer arrived, and what made them look. Then review where the two disagree.

1. Record both on every lead. The tracked source comes from tagged links (often called UTM tags) saved in hidden form fields, which does UTM tracking get blocked by ad blockers shows how to keep when analytics is blocked, and the self-reported source comes from the form and the first call.
2. Carry both onto the opportunity, so signed customers and their revenue can be counted by either field.
3. Each month, list the leads where the fields disagree. "Tracked: Google search, said: a referral" is the common case, and it usually means the referral created the demand and search captured it.
4. Credit the budget decision to what created the demand, and the campaign management to what captured it. Branded search still deserves its spend if it catches people who were referred, but it shouldn't be credited with the referral.
5. Watch the trend more than any single answer. One buyer may misremember, while a steady share of new customers naming the same association or podcast is a signal worth acting on.

When Does Self-Reported Attribution Mislead?

Self-reported attribution misleads when the answer is collected badly, when too few answers are read as a trend, or when it's used on its own to judge channels buyers rarely notice.

• Channels that work in the background. Display, video and retargeting ads build familiarity that buyers rarely credit, so self-reported answers will undercount them, and tracking plus a test (pausing the channel in some regions and comparing results) is the better judge there.
• "Google" as an answer. It often means "how I found your phone number", so on its own it says little about what sent the buyer searching.
• Long sales cycles. An answer collected at signing, months after the first contact, is more likely to name the most recent touch than the first.
• Buying committees. On a deal with several decision-makers, the person filling in the form may not be the one who first heard about you, so ask on calls with the others too.
• Small numbers. A handful of answers naming a new channel is a lead to investigate, and too few to move budget on yet.
• Polite answers. Buyers sometimes name a salesperson or a well-known source because it seems like the expected answer, which a follow-up question on the call usually sorts out.

None of these is a reason to drop the question, because tracking has its own blind spots, and each source corrects the other only if you keep both.