Case Study: Target Sales Analysis |
|||||||||||||||||
|
Retailers experience a great deal of demand-side seasonality. Here is an example of how Performance Chain™ can analyze sales for Target department stores. Target, like other department store retailers, has a long
history of consistent sales growth. Seasonality is clearly evident, as
well. Target’s fiscal year begins in
February, and consists of 4- and 5-week periods, in a When we break out the seasonal component, we see that the seasonality has been getting less severe. The strong holiday sales period has declined from being more than 80% above average to around 70%, and the low post-holiday low has improved from being somewhat more than 20% below average to being somewhat under 20% below. Perhaps this is the result of more aggressive post-holiday sales? Here is a close-up of the monthly seasonality. Note how March, June, September, and
December all spike up due to the extra week.
December is especially large, however, and September is somewhat
small. Performance Chain™ automatically
adjusts for the Random variation from month to month tends to be +/- 2-4%. This is relatively small, and will increase as you drill down to smaller, more-detailed levels. That leaves the trend, and we see it has indeed been growing smoothly during this time frame. Note that there is no seasonal pattern that repeats each year. All seasonality has been removed. When we look at the percent change from the previous period, we do see some variation in growth. While some of this may be random, there are dips during the recessions in the early 1990s and in the late 1990s/early 2000s. Performance Chain™ provides this breakdown for any level of detail, down to a department within a store. It makes sophisticated sales analysis practical for managers at all levels. Forecasting is also easy: · Seasonality is automatically forecast by Performance Chain™. · Random variation doesn’t need to be forecast – it is unknown, and averages zero over time. Performance Chain™ evaluates variances from forecast to determine if they are in fact random – if not, that is a sign that the forecast may be incorrect. · Performance Chain™ measures and segregates the effects of events as they occur. Users can easily forecast any planned events. · That leaves the trend. Performance Chain™ creates a default forecast of the trend based on its current level and growth characteristics. Users can easily incorporate their information and expectations and use their judgment to modify the default forecast. Because all other factors are segregated, it is very easy to evaluate and project the trend. |
||||||||||||||||
|
|||||||||||||||||
|
|||||||||||||||||
|
|||||||||||||||||
|
|||||||||||||||||
|
|||||||||||||||||
|
|||||||||||||||||
|
|||||||||||||||||
|
|||||||||||||||||
© 2004 Thinking Links, LLC All rights reserved. |
|||||||||||||||||