Self-Reported Attribution: How to Ask and Use It

Self-reported attribution means asking buyers directly how they found you, usually with a "How did you hear about us?" field on a form, booking page or sales call, and saving that answer on the contact and the deal in your CRM. It captures sources that pixels and UTMs cannot see, such as podcasts, word of mouth, WhatsApp forwards, events and creators. It works best as a second opinion next to software attribution, not as a replacement for it.

This guide covers how to word the question, where to place it, how to clean the answers, and how to compare them with tracked data without fooling yourself.

What is self-reported attribution?

Software attribution reads digital traces: click IDs, UTM parameters, referrers and cookies or cookieless identifiers. Self-reported attribution reads memory. You ask the person who converted, and they tell you what they remember as the reason they came.

The two methods answer slightly different questions. Tracking tells you which touchpoints happened before the conversion. The buyer tells you which touchpoint felt important. When both agree, you can trust the channel more. When they disagree, you have found a gap in your measurement worth investigating.

If you are new to the wider topic, start with our guide on what marketing attribution is and which models exist.

What does it catch that tracking misses?

Most of the value sits in channels that leave no click. A link shared in a private chat, a podcast mention or a conference talk usually shows up in analytics as "Direct" or as branded search. The table shows the typical pattern.

These hidden paths are covered in more depth in our posts on dark social attribution and measuring podcasts and other hard-to-track channels.

How should you word the question?

The wording decides what you measure. "How did you first hear about us?" measures discovery. "What made you reach out today?" measures the trigger. Fairing, a post-purchase survey vendor, describes four variations in its guide to question variations, each aimed at a different moment in the journey.

Pick one primary question and keep it stable for at least a quarter. If you change wording every month, you cannot compare periods.

How do you set up a "How did you hear about us?" field?

1. Choose the format. A dropdown with 8 to 12 options plus "Other (please specify)" gives clean data. Open text gives richer answers but needs manual tagging. Low-volume B2B teams can start with open text; teams with hundreds of leads a month usually need a dropdown.
2. Write options buyers recognize. Use "Instagram or Facebook", "Google search", "Podcast", "Recommendation from a friend or colleague", "Event", "LinkedIn", "YouTube", "TikTok", "Creator or influencer", "Other". Avoid internal jargon like "paid social".
3. Randomize the order if your form tool allows it. People pick the first plausible option more often. Keep "Other" last.
4. Place it where intent is high. Good spots are the demo or quote form, the booking confirmation page, the post-purchase page and the first sales call. For WhatsApp leads, the sales rep can ask in the first conversation and log the answer.
5. Decide required or optional with a test. A required field gives more answers but can cost form completions. Run it both ways for a few weeks and compare completion rates before deciding.
6. Store it as a CRM property on both contact and deal. In HubSpot, Salesforce or Kommo, create a dedicated field. Copy it to the deal so you can report revenue by answer, not just leads.
7. Normalize answers into channel buckets. Map free text like "my boss", "a friend" and "referral" into one bucket. Review the "Other" answers monthly to spot new channels.
8. Keep tracked source fields next to it. Capture UTMs and click IDs in hidden fields too, as explained in our guide to passing UTM parameters into your CRM. Without both, there is nothing to compare.

How do you combine self-reported and software attribution?

Build a simple matrix: tracked source in rows, self-reported answer in columns, closed revenue in the cells. If you already use an attribution tool such as Admira, the tracked source per deal is there; you only add the self-reported property next to it. Then read the disagreements.

Example (hypothetical numbers): a B2B services company closes 120 deals in a quarter worth $360,000. Tracking assigns 48 deals to Google Ads, 30 to Meta, 36 to Direct and 6 to referral sites. Self-reported answers for the same 120 deals say 22 came from a podcast, 31 from recommendations, 34 from Instagram or Facebook, 25 from Google and 8 from events.

In this example the podcast never appears in tracking, yet buyers name it in 22 of 120 deals. Cutting it because "it shows zero conversions" would be a mistake. Google Ads still matters, but part of its credit is really demand captured, not demand created. That is exactly the blind spot we describe in what last-click attribution hides.

A practical rule: use software attribution for daily and weekly optimization inside each ad platform, and use self-reported data quarterly to decide budget between channels that tracking undercounts.

What are the limits of self-reported attribution?

• Memory is short and selective. People tend to name the most recent or most memorable touch, not the first one.
• "Google" is ambiguous. Many buyers who heard about you elsewhere answer "Google" because that is where they typed your name.
• Option order and wording bias answers. A list that starts with "Instagram" will get more "Instagram".
• One answer hides a multi-touch journey. B2B buyers often see a brand through several channels before reaching out.
• Response rates vary. If only 40% answer, check that the answering group is not very different from the rest.
• Small samples swing. With 30 deals a quarter, two answers can move a channel by several points. Read trends over multiple quarters.

None of these make the data useless. They mean you treat it as directional evidence and pair it with tracked journeys or, for big budget questions, with incrementality tests.

Who this is for, and who it is not for

Good fit: lead-gen and B2B teams with sales conversations, brands investing in podcasts, events, creators or community, and companies in markets like Mexico and LATAM where a lot of sharing happens in WhatsApp groups.

Weaker fit: very high-volume, low-ticket ecommerce where a single paid channel drives almost everything and the extra form field costs more conversions than the insight is worth. In that case a short optional post-purchase survey is usually enough.

Where does Admira fit?

Admira connects Meta, Google Ads, TikTok and Click-to-WhatsApp leads to CRM revenue and shows the tracked buyer journey for each deal, so you can put a buyer's answer next to what actually happened. It does not replace the question itself: you still need the field in your form or CRM. If you mainly need lead-level source tracking, see Admira lead tracking; for longer B2B cycles, see B2B marketing attribution.

FAQ

Is self-reported attribution accurate?

It is accurate about what buyers remember, not about every touchpoint. Treat it as directional and compare it with tracked data rather than trusting either source alone.

Should the "How did you hear about us?" field be required?

It depends on your form. Required fields get more answers but can lower completion. Test both versions for a few weeks and compare completion rates and answer quality.

Dropdown or open text?

Dropdowns with an "Other" option give clean, reportable data. Open text gives richer detail but needs manual tagging, so it suits low-volume teams.

Where is the best place to ask?

Ask where intent is highest: the demo or quote form, the booking confirmation, the post-purchase page or the first sales call.

How many answers do I need before acting?

There is no fixed number, but a quarter of data and at least a few dozen closed deals per major channel make trends easier to trust than a single month.

Can self-reported attribution replace attribution software?

No. It cannot optimize campaigns daily or feed conversions back to ad platforms. It complements software by revealing channels that tracking undercounts.