Skip to content
mailprompt
Start free trial
GlossaryUpdated

What is email attribution?

Email attribution is the rule that assigns credit for an order to an email, a click or another touchpoint. The rule has to name the signal, the window and the last or first touch. Mailprompt uses a click-based, five-day, last-click model and reports opens without using them for revenue credit.

Key facts

  • Google Analytics describes attribution as assigning credit for a conversion to the touchpoints that contributed to it. Source
  • Apple says Mail Privacy Protection can prefetch remote content, which affects how email opens are measured. Source
  • Mailprompt credits revenue on a click with a five-day, last-click window and does not use opens for attribution. Source

What attribution means

This definition is part of the Mailprompt glossary. Attribution is the rule that decides which marketing touchpoint gets credit for a conversion. In email marketing, the conversion is usually an order and the touchpoint is usually an email, a click or a campaign.

The rule is not a fact about human behaviour. It is a convention that lets a team compare messages and periods. Two teams can look at the same customer journey and produce different revenue numbers because they chose different rules.

The four parts of a rule

A useful attribution rule names four things: the signal, the window, the model and the treatment of refunds.

The signal is what counts as a touch. A click is stronger than an open because it is a deliberate action. The window is how long the touch remains eligible. A five-day window is shorter than a thirty-day window and will credit fewer orders. The model is first click, last click, linear or something else. The refund rule decides what happens when an order is cancelled or returned.

If a report does not state those four things, the number is hard to interpret. A revenue total without a window is not a comparable metric.

Last click and its bias

Last click gives credit to the most recent eligible touch. It is popular because it is easy to explain and easy to compute. It also favours the message closest to the order, which is often a cart reminder rather than the newsletter that introduced the product.

That bias is not a reason to abandon the model. It is a reason to read it carefully. A welcome email may look weak on last click even when it creates the interest that a later flow converts. A team that only optimises for last click can underinvest in the top of the funnel.

Opens and the measurement problem

Open tracking works by loading a small image. Apple's Mail Privacy Protection can prefetch remote content, and security scanners do the same. The image may load before a person reads the message, or without the person reading it at all.

That makes opens a noisy signal. They are still useful for subject-line tests, where the comparison is relative and the noise is similar across variants. They are a poor basis for revenue credit, because a prefetch is not a purchase intent.

What to report

A report should show the signal, the window and the model next to the revenue number. It should also show the denominator: sends, clicks and orders. A revenue total without a send count can make a small flow look more important than it is.

Mailprompt reports revenue per email, per flow and per campaign on a click-based, five-day, last-click model. The number is an operating measure, not an accounting ledger. Treat it as a way to compare messages, not as a replacement for the store's finance report.

Frequently asked questions

What is last-click attribution?

Last-click gives credit to the most recent tracked touchpoint before the conversion. In email, that is usually the most recent clicked link. It is simple and explainable, but it favours the message closest to the order. A first-click model would tell a different story, and a multi-touch model would tell a third.

Why not attribute revenue to opens?

An open is a weak signal. Apple can prefetch remote content, security scanners load images and a person can open an email without buying. A click is a deliberate action that leads somewhere. Opens are still reported for subject-line testing, but they do not earn revenue credit.

Does attribution prove that the email caused the order?

No. It records a path that was observed, not every influence on the customer. A customer may see an email, ignore it, search for the brand and buy. If the search click is not tracked, the email gets no credit. Attribution is an operating number, not a proof of causation.

Sources