Spectra Trade Areas are probabilistic, meaning that models are applied to calculate the probability for households in a dissemination area to shop at any retailer. That probability is applied to metrics such as households, population, and dollars and buyers, which are further summed into reports. Understanding the methodology is key to knowing how to interpret the weighted metrics.
In this example of a grocery store trade area, the shading indicates the probability the households in the dissemination area will shop at that store. Some factors driving this probability include:
- Drive time and distance
- Size and channel of the store
- Competitors
- Urbanity
At the edges of this trade area, households likely have closer retailers and may occasionally shop at the focus store. The light shading indicates a lower probability, less than 4%. Any metrics calculated on these households will include a lower share of household, dollars or volume. In these light-shaded areas, 100 households are contributing less than 4 to the total household count.
Closer to the store, the shading is darker, indicating a higher probability of shopping at the focus store, greater than 15%. The demographics and shopping habits of these households will be weighted higher than the light purple area. 100 households in these dissemination areas will count as at least 15 households to the total household count.
Implications and Applications
The weighted Homescan metrics quantify the potential sales volume available for the retailer based on Homescan profiles. Exploring the gap or proportional gap (comparing the percentage of actual sales to the percent of potential sales) can highlight areas of strength and potential for improvement for a retailer. There may be situations where the full count of a demographic or households may be needed (i.e. determining the household count for flyer delivery) rather than the weighted count. In that situation, use the Data Export tool using the Spectra store trade area of interest and selecting the relevant demographic variable(s).