Better measurement · 4 min read
More email revenue is only half the story.
An attribution report tells you which purchases were associated with a message. It does not, by itself, tell you how many purchases the message caused—or whether they were profitable.
Begin with contribution per order
For a planning model, multiply net average order value by contribution margin. Use the same net-revenue definition for both, and subtract variable costs such as products, payment processing, fulfillment, and shipping subsidies. Do not deduct discounts or refunds twice. A $80 order at a 50% contribution margin leaves $40 before program costs, fixed overhead, and taxes.
Count the whole program cost
Include the service fee, additional email or data charges, and setup costs. In our illustrative $499 monthly pilot, $40 contribution per incremental order requires 13 additional orders to cover the monthly service fee alone. Including $250 per month of the $500 setup raises that to 19 orders. These are arithmetic scenarios, not forecasts.
Separate attribution from incrementality
A customer may have purchased without the email. Where practical, randomly assign eligible customers to a treatment group and a holdout group. Compare purchase outcomes over the same window, while accounting for group sizes. Decide the design and exclusions before launching.
Give uncertainty a place in the report
Small audiences and short windows may not produce a reliable estimate. Report that limitation instead of presenting a noisy result as proof. Watch for seasonal changes, promotions, changes in product availability, and overlapping campaigns.
Make a decision the business can use
Translate the evidence into a next step: continue the program, change the offer or timing, test longer, or stop. A program earns its place when its contribution and customer impact justify its cost.
Put it in the context of your store.
Request a free initial audit to find the next useful improvement.
Get my free audit ↗