Overview
One important aspect for any retailer to consider is the newness in their assortment. This often requires launching numerous products to locations that have never sold it before. This can often be a laborious task for retailers to figure out the expected performance of a new product in a location. To assist in this process, a location curve uses the history of similar products to index the locations against each other in order for demand to be spread down to a location level accurately and efficiently in the allocation process.
Too busy to read the article, watch this video on how to set up and apply a location curve.
Setting up a Location curve
Step 1: Create a Location Curve
To set up a location curve, navigate to Settings > Data Management > Location Curves and click on Add.
The following selections are available to generate a location curve
Location Curve Name - It is advisable to name the location curve in a way that it is easily recognizable when selecting for the appropriate product or time period.
Dynamic vs Static - Location curves can be specified as a static index entered directly by the user or as a dynamic curve which is calculated based on historical data.
Time Frame - The time frame allows you to restrict the data used for the location indexing. For instance, locations may behave differently in summer vs winter and this can be captured in the time frame selected.
Filters - filters allow for curves to be generated for specific types of product that may experience different performance per location. Shirts may perform better in some locations while shoes perform better in others.
Update Cadence - The update cadence is used to keep the location curve fresh when using a relative time period like the last 52 weeks. If you do not want the location curve to be updated, set the update cadence to Manual
Metric to Use - Depending on the source selected, Gross Sales Units or Demand Units (to account for lost sales) can be selected.
After making your selections, click on the SAVE button.
Step 2: Review and adjust the location curve as needed
In this step, users have the ability to select the location curve created and review the output in the screen below by double clicking on it. The indexing of locations is based on the productivity of the choices in the filters selected for the time period selected. Users have the ability to override the productivity to adjust the index for factors that may not be captured in the historic data.
Step 3: Apply the location curve to an allocation strategy
In this step, you have the option of selecting an appropriate location curve per allocation strategy, you can copy and paste the location curve to affect multiple allocation strategies. The demand for the allocated product will be spread across the locations as a result. Not the contribution of the top location vs the lowest contributing location.
Blended Location Curves
Location curve blending provides users with the capability to blend perspectives on how a new product may perform when allocating. The goal is to take more than one relevant data set into account when estimating the expected performance of a product.
Some examples of location curve blending are
When launching a new season, evaluate similar products in the previous year, same season. But, blend that with the latest trends of those products over the preceding X periods of time.
When evaluating a product that may not fit a single mold, for instance evaluate the footwear category performance but blend that with high priced footwear only to understand the unique characteristics of price point on demand.
Use a unique product as the basis for evaluation but blend that with a similar category due to the potential lack of information on the unique product.
Applying Location Curve Blending
When adding a location curve to an allocation strategy, users can simply follow the normal procedure of selecting the location curve from a drop down.
To add a second curve, double click or open the allocation strategy and navigate to the location curve tab. Click Add and select another location curve and the weighting to use for the curve.
Once the curve is added, adjust the weightings as required to hit 100%, you will not be able to save without the weighting equalling 100%
The blended location curve will now be used when spreading demand to the store level.