Is It Okay to Have More Than One Geography Selected in the CMA?

Geographies should be combined with caution, as numbers could get doubled if geographies overlap, and only basic measures will be calculated correctly with any combined geographies.

It’s important to tread carefully when combining geographies. Counties/stores can overlap and double up numbers, depending on which geographies are being combined. For instance, you wouldn't want to combine Tennessee-Food with Tennessee-Multi Outlet, because Multi Outlet geographies include food/grocery retailer data. In contrast, Tennessee-Multi Outlet could potentially be combined with Tennessee-Convenience, because "Multi Outlet" geographies don't include convenience or liquor retailers. 

Additionally, any time that geographies are combined, weighted measures like CWD (Category Weighted Distribution) will not calculate correctly. If there’s no geographic or store selection overlap, basic measures involving dollar sales and volume sales will calculate correctly, but Circana’s CWD calculation will not work across multiple geographies/projection models. Circana data will just display the CWD for whichever geography had the highest CWD in the workbook's timeframe. Causal measures involving Features, Ads, and Displays will not calculate correctly with combined geographies either.