A store contains products that play different roles. Some attract visitors, others complete a purchase, or respond to a frequent need. Comparing them only through turnover can hide the contribution they offer to the assortment.
I propose to observe margin, rotation and relationship with other products. An undersold reference could be important to make the service complete; a much-requested item could absorb space and work without contributing enough to the costs. These possibilities require data and observation, not labels assigned once and for all.
In a hypothetical example, an accessory generates purchases along with a main product. Evaluating it in isolation may overlook that relationship. However, it is necessary to distinguish observed pairings from simple expectations and to check whether the same space can be used better.
The revision of the assortment should therefore link the commercial function and cost of the resources committed. Reducing categories can simplify the store, but also leave needs uncovered. Increasing them can enrich the supply, requiring more capital and skills. The proposed criterion is to verify how each category contributes to the customer experience and to the economic sustainability of the whole, keeping the objectives of the choice clear.