In terms of price management, many brands are used to categorizing products according to their nomenclatures and the department where they will be sold. But this referencing logic alone is too rigid to allow fine pricing to be achieved, therefore to guarantee good profitability.
With data science, it is now possible to create more diverse product categories, based not on the description of the products but on the perception that customers have of their value, with the aim of offering prices adapted to their expectations and improve the performance of retailers.
Once the matrices used to create them have been developed, we can identify optimized pricing scenarios. It is then a question of simulating and analyzing their ROI in order to select the best. To achieve this, it is therefore crucial to define relevant categories. Here are three examples.
“Engine” and “additional” products
The analysis of receipts makes it possible to identify which products trigger the consumer’s visit to the store (whether physical or online). They are sometimes bought alone and sometimes with others: these are the “motor” products.
Other products are never purchased alone but always with one or more others, sometimes in connection with their use, sometimes not. These are the “additional” products that consumers take advantage of while they are in the store to buy, but which do not trigger their arrival.
Finally, we identify “independent” products, always purchased alone. To classify products into these categories, a systematic and large-scale analysis of receipts is carried out. This makes it possible to calculate the frequency of purchase “alone” or “in group” of each individual product, and therefore to classify them either in closed categories or on a scale.
The baskets-type is an approach closely linked to customer behavior. It consists in identifying, again on the basis of sales receipts, recurring standard baskets which make it possible to isolate distinct customer profiles. These profiles can then be refined by integrating, for example, data from loyalty programs or filtered by store.
In specialist distribution, this may correspond to specific projects rather than customer archetypes (bathroom redesign, furniture construction, etc.). Many consumers do not know precisely the price of each individual product but have a fairly clear idea of the overall price of their regular baskets, so it is essential for distributors to be able to compare and position themselves on this scale.
Customer’s sensitivity to price
This categorization corresponds to strategies focused on the image-price, that is to say on the perception that customers have of the relationship between the value, usefulness, satisfaction that a product brings them and its price. It is a question of categorizing the products according to this sensitivity.
This perception depends on a multitude of objective and subjective factors and varies over time, so the challenge here is not only to rely on data as rich as possible to define the right price, but also to make it evolve with it. time.
The finesse and precision of these categorizations strongly depend on the ability of retailers to accumulate and analyze increasingly vast fields of data and stemming from increasingly varied sources. The volumes reached today are such that specialized technologies for data science or the implementation of artificial intelligence appear as the only way to carry out these analyzes quickly enough for them to produce relevant operational results.
It is not only a question of defining a price, but of putting it in the context of the customer to whom it is addressed, of the store where it is sold, of current trends, of the brand’s brand image. and its profitability targets.
Georges Bory is co-founder of ActiveViam, publisher of BtoC pricing solutions.
Expert opinions are published under the full responsibility of their authors and in no way commit the editorial staff of L’Usine Digitale.