Foundation · 2 min
What you need to know
Should ecommerce metrics be compared day over day?
Day-over-day comparison is useful for immediate operational monitoring but often too noisy for business conclusions. Compare equivalent weekdays and longer completed periods for decisions.
Week over week vs year over year: which should I use?
Week-over-week highlights recent movement; year-over-year better controls for annual seasonality. Use both when available and annotate calendar, promotion and assortment differences.
How does seasonality affect ecommerce analytics?
Seasonality changes demand, product mix, conversion, order value, acquisition cost and returns at predictable times. Compare equivalent seasonal windows and keep cohort ages aligned.
Why can yesterday’s data be misleading?
Yesterday may contain late events, unsettled payments, unprocessed cancellations or refunds, reporting delay and weekday noise. Label it provisional until the normal completion window passes.
How do promotions distort normal performance?
Promotions change traffic intent, prices, product mix, timing and margin, so promotional performance is not a normal baseline. Report incremental contribution and post-promotion effects separately.
How should major sales events be compared?
Compare major events with equivalent event days, matched weekday and season, and a non-event baseline. Adjust for event length, promotion depth, inventory, channel spend and late returns.
A quick store check
Compare the same completed time windows and annotate promotions, outages, holidays and tracking changes.
Foundation · 2 min
Frequently asked questions
Should incomplete current-day data appear on reports?
It may appear for operations, but label it provisional and do not compare it with completed periods as if equivalent.
Can year-over-year comparisons remove every seasonal effect?
No. Holiday timing, promotions, assortment, prices and external conditions can still differ.