Attribution Models

Attribution answers the question: "How should this conversion be credited?"

When a user sees your ad on Monday, visits your site on Tuesday, connects their wallet on Wednesday, and makes their first swap on Thursday—did the ad drive that conversion? Attribution models provide a consistent framework for connecting exposure to outcomes.

Important: Attribution does not prove causality. It establishes a measurement system so performance can be tracked and compared.

Last-Touch Attribution

Credits the final marketing touchpoint before conversion.

Example: User sees your ad, clicks, connects wallet, and swaps. The ad gets full credit.

Simple to implement, clear credit assignment, realistic for web3 where journey visibility is limited. However, it ignores earlier touchpoints that built awareness.

First-Touch Attribution

Credits the initial marketing touchpoint.

Example: User saw your ad last week, later saw you mentioned on Twitter, then converted. The ad gets credit, not Twitter.

Values awareness-building and works well for long consideration cycles. May overvalue touchpoints that didn't actually drive the conversion decision.

Multi-Touch Attribution (MTA)

Distributes credit across multiple touchpoints.

Example: User sees ad (30% credit), reads blog post (30% credit), sees retargeting ad (40% credit), then converts.

Provides a more holistic view but requires tracking the full customer journey. This was best practice in web2's cookie-tracking era when user journeys were fully observable. That world no longer exists.

In today's privacy-focused environment, you can't see all touchpoints—especially with web3 users who value privacy and are cautious about clicking links. A common web3 pattern: see an advertisement, then convert hours later from a trusted link to avoid phishing scams. MTA will incorrectly over-attribute to that final trusted touchpoint.

There's no perfect model. The ideal approach varies by project, data infrastructure, and business goals.