“How did you hear about us?” — the one survey question worth more than your pixel

2026-07-07 · Customer Luv Surveys

Meta says the sale came from Meta. Google says Google. Your pixel lost the trail two devices ago. The customer, five seconds after buying, will just tell you — if you ask.

Why platform attribution over-credits itself

Ad platforms grade their own homework: view-through windows, modeled conversions, and last-click bias all inflate the channels that are easiest to measure. Meanwhile podcasts, retail shelves, word of mouth, and creator mentions — often your best-converting discovery moments — report almost nothing. Survey-based attribution consistently reveals 20–40% of revenue coming from "invisible" channels.

How to ask it properly

• Ask on the thank-you page. Response rates there are dramatically higher than email surveys sent days later — the purchase is seconds old and goodwill is at its peak.
• Randomize the options. Fixed lists over-collect the first option.
• Match options to your actual channels — including offline ones: "Saw it in a store", "Friend or family", "Podcast".
• Always include Other + text box. The write-ins are where you discover channels you didn't know existed.
• Ask "where did you FIRST hear about us" if you care about discovery rather than last touch.

Turning answers into budget decisions

Sync answers to Klaviyo as profile properties and the data starts working: build a segment of "survey_source = Podcast" and check its LTV against your Meta cohort. Compare claimed-source revenue share against platform-reported share — channels that customers claim but platforms don't are underfunded; the reverse are over-credited. Rebalance quarterly.

What good looks like

A healthy setup: attribution question first, randomized, skippable; three to four targeted follow-ups; a gift at the end. Expect 30–50% of buyers to answer the first question when it's asked on-page — versus low single digits for emailed surveys.