Reading your own numbers
Which figures tell you something, which mislead, and how to compare periods honestly.
Most shop dashboards show a lot of numbers. Three or four of them decide whether the business works.
The ones that matter
Orders and average order value — what people buy and how much at a time.
Conversion — of the people who arrived, how many bought. A fall here is usually something broken, not something about the market.
Returns and refusals — the quiet margin eater. Revenue before returns is not revenue.
Comparing periods
Compare like with like: the same days of the week, the same length. A Monday-to-Sunday week against a Tuesday-to-Monday one will show you a difference that is not there.
Be careful comparing against a peak. A normal week will always look bad next to a sale week, and reacting to that is how good campaigns get switched off.
Where the truth is
Your own orders. Figures from an advertising platform count what it can see and attribute generously; figures from a pixel miss what blockers hide. Both are directions. Orders are the number.
What to check
- The revenue figure matches what arrived in your account over the same period.
- You know whether the figure is before or after returns, and before or after delivery.
- A number you are about to act on exists in two places, and agrees.