8 Ways to Ask 'How Did You Hear About Us?'

The "How Did You Hear About Us?" question is the cheapest attribution data you own. Unlike pixels that break when iOS updates ship, and unlike UTM parameters that vanish in private browsing, a well-designed HDYHAU question gives you a direct line from customer to channel — in their own words.

The problem isn't whether to ask. The problem is how. Most ecommerce brands pick a single question format, configure it once, and collect years of data shaped more by survey mechanics than by customer reality. A static dropdown inflates the first answer by default. A missing option sends real channel data into an "Other" black hole. A vague time frame conflates the channel that first made someone aware with the one that pushed them to finally buy.

The good news: question design is a lever, and it's easy to move. Here are 8 question formats — each one fixes a specific data problem — and a guide for picking the right one.

Why the Default HDYHAU Question Falls Short

The standard version — a static dropdown with your main channels — fails in three predictable ways.

Primacy bias. In self-administered web surveys, respondents favor options at the top of a list. Research published in Public Opinion Quarterly found that in visual surveys, earlier-listed options are chosen at higher rates — not because they're more accurate, but because they're seen first. If "Google" or "Instagram" sits at the top of your list, those channels are inflated by design.

Recency bias in memory. Customers remember the last channel they touched before buying, not the one that first made them aware. A shopper who heard about your brand on a podcast eight months ago and saw a retargeting ad yesterday will almost always credit the retargeting ad. Without a time dimension, you're capturing conversion triggers — not always true acquisition channels.

Missing options. If your list excludes "a friend texted me," "a Reddit thread," or "a newsletter I subscribe to," that customer has no honest answer. They'll select whatever feels closest or abandon the question. The data looks complete. It isn't.

Formats 1–4: Fix the Structure First

These changes affect how the question is presented. They don't require rewriting the question — just rethinking the mechanics.

Format 1: Randomize the Answer Order

The simplest upgrade. Randomize channel options on every survey load. This distributes primacy bias evenly across your answer set so that your aggregate reflects actual attribution rather than list position. SurveyMonkey's research on order bias confirms this is standard practice in reliable survey research — and it's a one-toggle fix in most tools.

Best for: Stores with 500+ monthly orders, where statistical averaging relies on randomization to work accurately.

Format 2: Forced Choice Instead of "Select All That Apply"

Asking customers to select all applicable channels tends to produce inflated, low-signal responses. Respondents check boxes without prioritizing. Pew Research Center's comparative study on forced-choice versus select-all formats found that forced-choice questions produce more deliberate, higher-quality answers.

Instead of "Which channels influenced your purchase? (Select all that apply)," ask: "Which single channel most influenced your decision to buy?" You lose a layer of nuance. You gain clarity.

Best for: Multi-channel advertisers running spend on 3+ platforms who need a clear signal for budget allocation.

Format 3: Add a Time Dimension

Add a follow-up: "About how long ago did you first discover us?" Options: within the past week / 1–4 weeks ago / 1–6 months ago / over 6 months ago.

Many customers follow a brand for months before buying. A shopper who found you through TikTok content nine months ago and clicked a Google remarketing ad yesterday will credit Google — unless you ask when. The time frame separates your acquisition channel from your conversion trigger. Both matter. They're not the same thing.

Best for: Brands with longer consideration cycles — supplements, fashion, higher-ticket items — where the gap between awareness and purchase is wide.

Format 4: Open-Ended First, Closed List for Confirmation

Instead of jumping straight to a dropdown, ask first: "In your own words, how did you first discover us?" Then follow up with a categorized list to classify the response.

Open-ended answers surface channels that never occurred to you to include. Picture a brand that discovers 12% of new customers mention a regional podcast or a specific online community — neither of which appears on their standard list. The tradeoff: slightly lower completion rates and more manual analysis.

Best for: Brands in the first 12 months of growth, or any store entering a new market where channel discovery matters more than data volume.

Formats 5–8: Go Deeper on Signal

These formats extract richer data by asking more specifically — splitting the question, following up on key answers, or capturing channels that closed lists miss entirely.

Format 5: Split Awareness from Intent

Ask two questions instead of one:

1. "How did you first hear about us?"
2. "What made you decide to buy today?"

These answers often diverge. The first maps acquisition channels — where new customers originate. The second maps conversion triggers — what finally pushed them over the line: a discount, a review, a friend's message. Conflating the two produces data that looks like attribution but isn't.

"The channel that first made someone aware of your brand and the channel that made them buy are often two different things. Ask once, and you tend to capture only the second."

Best for: Any brand investing in both awareness content (podcast ads, influencer posts) and bottom-of-funnel retargeting simultaneously.

Format 6: Referral-Specific Follow-Up

When a customer selects "Friend or family recommendation," add: "How did your friend share it with you? (Text/message, in person, social media post, other)"

This matters because word-of-mouth travels through channels that are invisible to every pixel and UTM parameter. Nielsen's Trust in Advertising research found that 88% of consumers trust recommendations from people they know above all other forms of advertising. But that trust travels through private messages — WhatsApp, iMessage, Slack — which analytics misattributes as "direct" traffic by default.

Research on dark social traffic estimates that 100% of clicks from private messaging apps go unattributed in standard analytics. A referral follow-up question begins to make that invisible channel visible.

Best for: Beauty, health, supplements, and community-driven DTC brands where WOM referrals are likely already large and uncounted.

Format 7: Channel Plus Content

Extend your question to content-level data: "Where did you discover us? And if you remember, what specifically caught your attention?" The second part is an optional open-text field.

This turns channel attribution into content attribution. "TikTok" becomes "a video about your refillable packaging." "Instagram" becomes "a comment on a skincare comparison post." That's a different class of insight — useful for creative teams and content strategists, not only media buyers.

Best for: Content-first brands with active organic video or social presence who want to connect specific creative to acquisition, not just channel to acquisition.

Format 8: An "Other" Field That Actually Works

If you include "Other" on your list — and you should — make it a required text field when selected. "Tell us where! We'd love to know." should appear the moment a customer clicks Other.

Over time, patterns emerge: "Reddit," "a newsletter," "a friend's podcast recommendation," "a pop-up store event." Review these responses monthly. When any single source appears more than a handful of times, promote it into your core list as a first-class option.

"The 'Other' field without a text box tells you a customer didn't fit your categories — and nothing else. The text box is what turns it into research."

Best for: Every store, as a default addition to any of the formats above.

Choosing the Right Format for Your Store

No single format works for every situation. A fast decision guide:

Your situation	Recommended formats
500+ orders/month, multi-channel ads	Format 1 (Randomized list) + Format 8 (Other text)
Under 200 orders/month	Format 4 (Open-ended first)
Supplements, fashion, longer purchase cycle	Format 3 (Time dimension) + Format 5 (Split question)
High WOM or referral brand	Format 6 (Referral follow-up)
Content-led or creator brand	Format 7 (Channel + content)
3+ active ad channels, budget to allocate	Format 2 (Forced choice)

For most stores, a randomized closed list with a working "Other" text field (Formats 1 + 8) is the right default. Add Format 5 if you run both awareness and retargeting. Add Format 3 if your customers tend to research for weeks or months before buying.

What to Do Once You Have Better HDYHAU Data

Better question design gets you higher-quality raw data. The next step is putting it to work alongside your pixel and UTM data — each layer sees something the others miss, and the gaps between them are where your real attribution blind spots live.

For a step-by-step process, see Reconcile Survey Data with Pixel and UTM (No Data Team). For a full breakdown of how each acquisition channel behaves in survey data versus pixel data, the Complete Guide to Channel Attribution Surveys covers it channel by channel.

Ready to start collecting attribution data your dashboard can't fake? Try Rauxdata free and add your first HDYHAU question in minutes.