Weekly Retail Calendar

Reporting is important part of every business. You need to understand how you performed on monthly, quarterly and yearly basis. In most cases you also compare the reported data with earlier similar reports, be it this month vs previous month, or last year's corresponding period.

When you make comparisons it is important to bear in mind that comparison needs to be comparable in all terms. Let's say you want to compare December 2019 and December 2020. Retailers make big chunk of the sales in weekends and there are four weekends in December 2019 and five weekend in December 2020. Or if you compared October 2021 with October 2022 it even more interesting as in October 2022 there's four weekend plus one Sunday. You want to make sure the variance reports are actually comparable.

This is why in is common to use weekly calendar in retail. This means that the fiscal year consists of 52 weeks (or 53 in every 4-5 years). As we saw earlier the weeks do not necessarily fit to common 12-month calendar there are numerous ways to make sense in all this, such as the 4-5-4 calendar (see https://nrf.com/resources/4-5-4-calendar).

ForecastingApp enables you to use any implementation of the weekly calendar, as the weeks are created for you based on the ISO 8601 standard (see https://en.wikipedia.org/wiki/ISO_week_date). For convenience we've created approximate 12-month calendar based on the standard. Basic rule of the standard is that if January 1st is Thursday or earlier then this entire week is part of the new year. We've applied the same logic to the months for convenience. As a result, all months have either 4 or 5 of 7-day weeks each. While this does not fit perfectly to monthly reporting cycle, it is always comparable, even if the previous year's same month does not have same number of weeks. Rather than comparing months we compare weeks and same number of weeks.

Another important aspect of having weekly calendar is forecasting. Often monthly forecasts are too high-level in order to support operational activities such as scheduling purchases or planning promotions. Days, on the other hand, are too detailed and the historical data is too volatile and contains too many peaks and gaps in order to produce precise results. Also, making daily forecasts for entire product range is time consuming for the perceived benefit.

So you create weekly forecast and you can be sure that each forecasted (or reported) period has exactly seven days and includes one weekend. Occasionally there are other events such as Christmas but these are easy to keep track on, and our inbuilt forecasting engine is aware of these.