An ecommerce scorecard should tell you what needs attention and which decision comes next. Start with revenue, spend, new customers, order contribution and purchase progression. Put the definition and reporting period beside each number. A dashboard with twenty unexplained ratios is harder to use than five numbers everyone understands.
A fictional 5.0 MER that does not prove profit
A fictional store records $100,000 of revenue and $20,000 of paid media spend in the same month. Using revenue divided by paid media spend, its media efficiency ratio is 5.0. The calculation is $100,000 / $20,000.
Now suppose the defined revenue leaves $32,000 after product and other variable order costs. Subtracting the $20,000 media spend leaves $12,000 before fixed overhead and other excluded costs. The MER did not reveal those costs. Another store with the same MER could have very different economics.
Label whether revenue includes discounts, refunds, tax and shipping. Specify which spending is included. MER is a useful aggregate ratio under a written definition, not proof that ads caused all the revenue or that another budget increase will pay back.
Choose rows with a job
For each row, record the current period, comparable prior period, definition, source and action question. A useful starter set is:
- Defined store revenue: did sales change, and were refunds or promotions involved?
- Paid media spend: are all relevant accounts and currencies included?
- New customers: does the available history distinguish first purchases reliably?
- Acquisition cost: is this media-only cost per new customer or a broader marketing-cost measure?
- Contribution: which variable costs have been deducted, and what remains excluded?
- Purchase progression: where do eligible sessions move toward an order or stop?
Add customer value on a fair timeline
A recent customer cohort has had less opportunity to buy again. Compare groups at the same age and show the customer count beside the percentage. Separate observed revenue within a defined window from a forecast of lifetime value.
Use the repeat-purchase audit to document the cohort, exclusions and cutoff. If the data is incomplete, retain the row with an explanation instead of filling it with a benchmark from an unrelated category.
Run the review as a decision meeting
Bring three notes: the largest meaningful change, the explanation supported by evidence and the question still unresolved. Assign one owner to the next investigation. A fall in conversion could coincide with a stockout, a new traffic mix or a tracking change; the scorecard identifies where to look, not which cause to assume.
Keep promotions, stock availability and reporting changes beside the period. Compare completed windows, and allow enough time for refunds and reporting delays where they matter. Your next step is to define one disputed metric in writing before adding another dashboard.
Editorial revision, October 20, 2026: unsupported category targets and reported client results have been removed. The numbers above are fictional arithmetic, not recommended operating thresholds.