For category management, many process models such as the famous 8 step approach or the day-to-day category management process have been defined. My experience, however, in discussions with many retailers worldwide, was that such process definitions can provide a good entry into the philosophy of category management, but ultimately, each retailing company has to find its' best process approach on its' own. Reason for that are: different strategies to address a market, heterogeneous consumer behaviour and cultures and existing IT landscapes and organizational structures within the company. As histroy can prove, several global acting retailers already failed in entering a market by just copying the successful process of the home country to the new one. In addition, retailers often underestimated the need for change management, to establish or reingeneer a complex and important process such as efficient assortment and space. In the following, I would like to briefly introduce a process of efficient assortment and space, which in my opinion, serves the process requirements for retailers pretty well.
Category Definition:
Similar to the 8 step approach or other best practise approaches as a basis for an efficient analytics and planning of a category a hierarchy has to be defined to determine which articles belong to the specific category. In order to put the cosumer into the focus of this process a consumer oriented hierarchy (Consumer Decision Tree - CDT) is created which reflects the decision process of a consumer when he wants to fullfil the specific need covered through the category. It is important to mention that all articles available in the market place belong to this hierarchy and not only the already listed ones. Outpunt of this process is a revision of the assortments and space plans for the relevant category. Due to that all potential articles have to be considered. The structure of a hierarchy either can be internally defined by the retailer himself or imported from data providers of syndicated data. Necessary functions have to be available to handle the articles belonging to the category which have not been listed so far and due to that are not yet known in the own system (so called market articles).
Category Analytics:
In this process step the category and its sub-structures are analyzed to get a clear understading about:
Market: How are the market trends and how do I perform against the market?
Competition: How do my main competitors perform in this specific category?
Supplier: How successful perform certain brands withing the category?
Consumer: How loyal are my consumers? How many % of all consumers do I address with the assortments of my category?
Main data sources to answer the questions above are POS Data, Market Research data and Space data. The consumer decision tree (CDT) mentioned in the previous process step is a key factor in category analytics in order to be able to analysis certain data such as POS data along the decision process of the consumer and understand certain patterns in behaviour.
Category Measurement:
In this process step all relevant goals are set which shall be achivied with the category or sub-category. Goals defined through different KPIs can either be directly related or an input from previous planning processes such as strategic or merchandise planning. The planning happens normally on category or sub-category level and contains financial, market, space and cosumer related KPIs.
Store Clustering:
Before an assortment (meaning a certain amount of articles out of the relevant category) can be planned, the retailer has to plan how much space per category he wants to assign to every store. This means that for one category several assortment moduls representing different sizes and used space (e.g. 3m, 4m and 5m shelf plan = planogram) have to be planned. The outcome of this process step are for the given category the different clusters with the relevant stores assigned to it. KPIs for clustering stores can be financial, demographical or also attributes of the store itself. Optionally the integration of a floor planning program can be conducted to assure that the assortment size really fits into the relevant stores due to building restrictions.
Assortment Planning:
Now per cluster the articles have to be planned which build the specific assortment for this category. Potentially all articles can be part of this assortment which are assigned to the CDT, meaning that some articles will be listed new, some will stay in an assortment, some will be discontinued and some new market articles will not be considered. In order to find out which articles should have a higher priority to become part of the assortment, one or several KPIs with weighing factors generating an index are determined. Through this one KPI or the generated index a ranking of the article is achieved showing the importance of the article within the category. Now either half automated by the generation of an assortment proposal or manually the articles are selected which shall belong to the category. How many articles are selected depends on different aspects such as space constraints, the importance of the specific category for the retailer (Category role) and the fact to generate a balanced assortment containig articles from all segments of the category. Dependencies within the different assortments of a category (e.g. all articles of the small assortment shall also be in the medium) can be considered if wanted. Simulations of what if scenarios in order to optimize the assortments accordingly are added value. The outcome of assortment planning is an assortment proposal per cluster for the relevant category. In order to verify more in detail that the chosen articles fit into the assortment and the determined space and to optimize the space used per article the assortment proposal can be communicated to a shelf optimization program where necessary adjustments can be carried out and communicated back to planning.
Category Operations:
After finalizing the process of assortment analytics and planning a key element of effecient assortment and space is to efficiently and flexibly communicate the results (meaning assortments, master data, store assignments, etc.) to the operational ERP system to ensure that the lans can have the immediate effect and follow up process such as listing, replenishment, etc. can be carried out in time. Here especially the space parameters such as placement order of an article within the shelf or shelf capacities can be of great value. Also more and more store specific planogramming is coming to the market where the assortment per store is further analysed based on POS data and store specific adjustments can be done. This normally directly happens in conjunction with the operational system because a store specific assortment planning would be too much workload in the processes above.
Category Implementation:
In order to implement the assortment changes in time in the stores an efficient communication line from the back office to the stores have to be established. New graphical shelf plans have to be sent out in order to implement the changes, the POS system needs the right data and prices and the printing of shelf labels should be provided.
Category Review:
Only if regurlarly the results of the plans are reviewed (mainly based on the KPIs defined in the process step Category Maesurement) both positive and negative signals can be realized in time and actions can be started. Another relevant aspect of this process step is to check the accuracy of the implementation of assortment changes in the store because only if the planned changes are really executed the planned goals can be achieved.
In Conclusion:
It is important to mention that not always the full process of efficient assortment and space has to be carried out. For example a CDT is not revised every year besides the assigned articles which change regularly. Different process steps are used in different periodicity. Whereas the planning steps are only done yearly or every half year if the overall categories are revied and newly planned, process steps like category analytics or category operations including space management are carried out more often due to ongoing article introductions during the year.
Finally I would like to emphasize that Category Management is a complex process containing of so many process steps that the requirements to this process can differ very much from one retailing company to another. Due to that an IT enabled category management solution can only be successful if it consists out of many highly flexible process components (or even services thinking about the Enterprise Service Oriented Architecture (Enterprise SOA) of SAP) to easily model the process with those components which are needed and to integrate maybe existing components into the process landscape accordingly.
