How did you hear about us? A survey guide with answer options

Updated 2026-11-02

"How did you hear about us?" may be the most useful single question in marketing. It takes minutes to add, costs nothing, and picks up channels your analytics can't see.

This guide covers where to ask, which answers to offer, and what to do with the results.

Why this question matters

Tracking sees clicks. It doesn't see the podcast someone heard while driving, the friend who mentioned you at dinner, or the AI assistant that recommended you. When those people arrive, they often type your URL, and your reports show them as Direct.

The answer is often the only record of that channel. Asking gives you data on dark social, word of mouth and offline channels that no script can reach. This is known as self-reported attribution.

It has weaknesses too. People forget, and they tend to name the most recent or most memorable thing. So treat it as a companion to tracking, with each one checking the other.

Where to ask

On the signup form. This is the best default. Everyone who signs up sees it, and the moment is fresh. Add it as one more field.

On the checkout or post-purchase page. This focuses your data on people who paid. A thank-you page is a low-risk spot, because the purchase is already done.

In your onboarding flow. You can ask inside the first session, right after signup, as a single question.

In a welcome email. A plain reply-to question often produces rich answers, with fewer people replying.

On your sales calls. Reps should ask and record the answer. Buyers talk more freely in conversation.

If you can only pick one, go with signup.

Should it be required?

Think about your volume. A required field gives you more answers and can reduce signups a little. An optional field keeps your form light, though fewer people answer.

A simple approach is to start with optional and watch your answer rate. If too few people respond, try making it required and see whether signups change. Measure the result on your own form.

Multiple choice or free text?

Each has a tradeoff.

Multiple choice is easy to answer and easy to report on. Its weakness is that you only hear about options you thought of.

Free text surfaces surprises. People will name the exact podcast or creator. It's also messy, and it takes work to group.

The best version combines them: a short list, an "Other" option, and a follow-up text box. If you use free text alone, you'll want a tool or a process that groups the answers.

Answer options to use

Here's a good starting list for a SaaS product. Adjust it to the channels you really use.

• Search engine (Google, Bing)
• AI assistant (ChatGPT, Claude, Perplexity)
• Podcast
• YouTube
• LinkedIn
• X
• Reddit
• Other social media
• Newsletter
• Friend or colleague
• Blog or article
• Event or conference
• Ad
• Other (please say where)

A few tips:

Randomize the order, if you can. People tend to pick items near the top.

Keep the list to around 8 to 14 options. Long lists slow people down.

Keep "Other" last. It should always have a text box.

List AI assistants as their own option. They've become a common way to find software, and they can be tricky to spot in referrer data.

Update it over time. If "Other" keeps turning up one source, promote that source to its own option.

Variations on the wording

The classic wording works. If you want a different angle, try these:

• "Where did you first hear about us?" This pulls answers toward first touch.
• "What made you sign up today?" This pulls answers toward the last trigger.
• "Who told you about us?" This suits referral-heavy products.

Pick one and keep it. If you change the wording, your trend will shift for reasons that have nothing to do with your marketing.

How to read the results

Review answers monthly. Here are the main things to look for.

Compare with tracking. Put the answers next to each person's first tracked visit. The gaps are the interesting part. Here's a made-up example of 200 signups:

Thirty-one of the 44 podcast listeners would've looked like Direct traffic without this question. Your analytics said nothing about the podcast, and your buyers did.

Follow the revenue. Count answers, then weigh them by what those people paid. A channel with fewer answers might bring much bigger accounts.

Watch the trend. Compare month to month. A channel that climbs after you start investing in it is a good sign.

Read the raw text. Exact words can name a creator, a community or a phrase you can use in your copy.

Mistakes to avoid

Offering too few options. If the real answer isn't there, people pick the closest one.

Treating answers as exact. It's a directional signal. Someone who answers "Google" may have searched your name after hearing a podcast.

Asking too late. Memory fades. Ask at signup, and not six months in.

Never grouping free text. A pile of raw answers gets ignored.

Collecting and not acting. Pick one thing to change each quarter based on what you find.

How to do this in Headstream

Self-reported attribution in Headstream takes minimal setup. Name your form field heard_about and the script reads it on submit, including dropdowns and radio buttons. You can also send answers through the identify call or the API, or import old ones.

Free-text answers go into groups automatically: Podcast, AI assistant, Search, YouTube, LinkedIn, X, Reddit, Other social, Newsletter, Friend or colleague, Blog or article, Event, Ad and Other. Custom rules regroup past answers too.

Each answer sits next to the person's first tracked visit, and you get a "hidden from tracking" stat that shows how much your analytics missed. If you've connected Stripe, which is in beta, you can see the revenue behind each answer.

For the full picture of how this fits with tracked data, read our guide to marketing attribution.