How we prove what the work earned.
Most reports show what happened after the work. We show what happened because of it, in your own data, with a method your finance team can check.
Why most reports flatter the work
Most tools use attribution, which gives a campaign credit for anything a customer did after receiving it. Many of those customers would have bought, returned or renewed anyway.
| Illustrative example | Spend per customer |
|---|---|
| Customers who received the offer | 40 |
| Similar customers who did not | 30 |
| What the offer actually added | 10 |
An attribution report would credit the offer with 40. The real effect was 10. Decisions taken on the first number spend budget on customers who never needed the push.
The method
- A holdout group, usually five to ten percent of customers picked at random, gets none of the new work for an agreed period. It shows what would have happened anyway.
- Each larger campaign also gets its own control group: customers picked at random who skip that one campaign.
- We count everyone who was picked, not only the people who opened a message, so engaged customers do not flatter the result.
- We measure net revenue over thirty days, after every cost including rewards and discounts, and check the weeks after for spend that only moved earlier.
- High value customers are spread evenly across the groups, so one large customer cannot decide the result by luck.
What you receive each month
| Monthly proof report | Source |
|---|---|
| Extra net revenue, compared with the untouched group | Your data |
| The cost of creating it: team, rewards, tools | Your data |
| Return on that cost | Calculated |
| Days for a new customer to pay back their acquisition cost | Before and after |
Your data, your numbers
We measure inside your systems, using your definitions, and agree the method with you in writing before anything starts. Every figure can be traced back to a saved query your team can run again. If a result is unclear, we say so rather than round it up.
Want to know what your current work is actually earning?
Bring one campaign to a thirty minute call. We will show you how we would measure it, and what the first report would look like.