Period comparison is the basis of every report, yet it is often done incorrectly. Comparison with the previous month is misleading whenever the business is seasonal or when months have a different number of working days and weekends. A year-on-year comparison of the same period tends to be more reliable, as it filters out seasonality, supplemented by a comparison with the previous period of the same length starting on the same day of the week. When evaluating, exceptional events must also be taken into account – a campaign, a website outage, a change in measurement, or public holidays – which on their own explain most deviations. A useful addition is a moving average over a longer period, which shows the real trend instead of the fluctuations of individual weeks. The report should also include a short commentary explaining significant deviations, otherwise every reader will interpret them their own way. This commentary is in fact the most valuable part of the entire report.
See also: Seasonality of demand, Annotations in analytics, Reporting cadence.
