Self-reported attribution
Updated 2026-11-02
Definition
Self-reported attribution means asking people how they found you and recording their answer. It's usually one field on a signup or checkout form: "How did you hear about us?"
Tracked attribution records where the last click came from. Self-reported attribution records what the person remembers as the reason. The two often disagree, and the gap is useful. People can name a podcast, a friend or an AI assistant's answer, none of which leave a clean trail in your analytics.
Example
This example uses made-up numbers. In one month, 200 people answer your signup question.
| They said | People | Tracking showed Direct or nothing |
|---|---|---|
| Podcast | 60 | 48 |
| Friend or colleague | 50 | 30 |
| Search | 70 | 5 |
| Other | 20 | 12 |
Tracking would have shown most podcast listeners as direct. Without the question, the podcast looks like it did almost nothing.
Strengths and limits
Self-reporting catches what tracking can't, including dark social. It also has flaws. People forget, pick the first option, or name the last place they saw you. Free-text answers need cleaning. Use it next to tracking and treat neither as the full truth.
How Headstream measures it
Headstream collects answers from a form field named heard_about, an identify call, the API or an import. It sorts answers into groups such as Podcast, AI assistant, Search and Friend or colleague. You can write rules that regroup past answers too.
Each answer is compared with the person's first tracked visit. The "hidden from tracking" stat counts people who named a real source that tracking saw as Direct or never saw. See self-reported attribution, or follow the setup docs.