--- language: en title: "Attribution gap" description: "The attribution gap separates potential brand touchpoints from what data collection and the model can attribute. Presence in AI does not equal a proven sale." canonical: "https://tropk.ai/en/resources/glossary/attribution-gap" last_modified: "2026-10-12T16:09:53.529Z" --- # Attribution gap Reviewed on 2026-10-07. ## What it means The attribution gap is the difference between touchpoints that may influence a decision and those that data and the attribution method can represent. In [AI search](/en/resources/glossary/ai-search), someone may discover a brand in an answer and reach the website another way. The possibility of influence does not automatically indicate its extent or allow that purchase to be attributed to the answer. ## Why it matters for marketing For marketing, recognizing this gap helps avoid evaluating AI presence solely through identified sessions or treating mentions as presumed revenue. Attribution models distribute credit according to available rules and information; they do not observe every external touchpoint. The team can track presence, traffic, and business outcomes as complementary evidence, keeping their units and limitations distinct rather than combining them into a metric without a method. ## Example in practice Imagine a manufacturer receives an order after a search for the brand name. In an interview, the buyer says they discovered the product in an AI answer, but the report records only the later search. The team keeps this account as complementary evidence about the same order, without automatically changing its source or concluding that all similar orders followed the same path. ## How to apply or verify Specify which interactions data collection observes, which model assigns credit, and what time window is used. Compare metrics using compatible segments and record reported touchpoints without confusing them with instrumented events. To investigate incremental effect, use a comparison suited to the question, taking other commercial activities and seasonality into account. When the data cannot distinguish causes, present the uncertainty alongside the result. ## Important distinctions The gap is not a universal rate or a value that can be recovered by multiplying assistant sessions. Missing referrers, no click, and credit rules are different issues. A customer statement reports their account; it does not prove the entire causal sequence. More citations can coexist with stable sales; fewer identified sessions do not, on their own, demonstrate less AI influence. ## Sources and references - [Get started with attribution]() — Google Analytics Help - [\[GA4\] Understand (direct) / (none) traffic]() — Google Analytics Help ## Continue a leitura - [Branded search growth]() - [AI referral traffic]() - [Traffic with no identifiable source]() ## About the author [Luiz Gardelin]() Luiz Gardelin is the founder of Tropk.ai and works in GEO/AEO, product vision, strategy, and growth, helping brands understand and improve their presence in AI responses. --- [All pages in Markdown]()