How does clustering help you find the right assortment in a category?
How does clustering help you find the right assortment in a category?
The demands of buyers in neighboring retail outlets can vary radically. Simply scaling the assortment to the size of the retail space is not an option. Such a solution can reduce customer loyalty and sales. How does clustering help to determine the needs of customers in each of the stores of a huge chain and fill the shelves with exactly the goods that they will buy? Read detailed analysis in our article.
What is clustering?
A well-chosen assortment in each product category is a key factor in increasing sales and profitability. To choose the right assortment, you need to study the history of sales, customer behavior and the demand of each individual retail outlet of the chain. It would be fair to argue that it is time-consuming and resource intensive. Clustering will help here. In the process of clustering, we identify patterns in consumption and collect stores into groups (clusters) with similar consumption.
For example, take the beer category. Let's break down the product categories according to certain properties:
- type of packaging
- country of origin
- price segments
- taste and varieties of beer
- imported or national brand
And we will try to find how the stores are similar to each other in the structure of consumption of a low-alcohol drink.
In mathematics, there have long been developed tools for cluster analysis. For example, clustering method and k-means search. We will locate the stores on the map of coordinates, calculate and analyze how the stores are similar to each other. The k-means cluster analysis method allows you to group outlets according to certain criteria.
For example, after analyzing, you can cluster store groups where:
- beer consumption of national brands prevails.
- prevailing consumption of craft or imported beer
- consumption of German beer in the low price segment is growing.
Clustering these stores allows assortment to be selected and optimized based on analysis rather than in an intuitive way.
Based on the obtained sample by groups, we will add beer sorts to the assortment of some outlets in accordance with the preferences of buyers.
Would you like to learn how Num8erz analytical tools work on case studies of large retailers and manufacturers? Watch the presentation on how Num8erz.InsightsPanel helps to analyze the switching of demand from a well-known juice brand and the launch of a new beverage. Click to download the presentation! If you have any questions, please write to [email protected] or fill out the consultation request form.
When developing clustering, it is worth paying attention to all the properties of goods, since even such a characteristic as the quantity in a package can be decisive when making a purchase decision. For example, out-of-town stores often sell large packs rather than small ones.
In order to understand exactly which goods are in demand in certain groups of stores of the trading network, a grouping is done. Clustering helps us in this. In the future, we saturate these groups with a profile assortment that is in demand in these outlets.
What should be considered in cluster analysis for retail?
For each category of goods, you need to do a different clustering of outlets. That is, clustering on the principle that expensive goods are sold here, or cheap ones are categorically not enough.
The Consulting for Retail (C4R) team studies each product category, conducts an in-depth study of clustering. Due to this, experts distinguish in each category those properties of goods that characterize the demand in a particular group of stores.
For example, based on data on beer consumption, a store can single out only three clusters, while pasta consumption can single out as many as five clusters. This is due to various factors:
- geographic location
- people who are ready to spend their money in our retail outlets
- income level of visitors
- competitive environment
Therefore, clustering must be carried out for each of the categories of goods, and it is necessary to group outlets in the context of each separate category of goods.
Do you want to calculate the Decision Tree or learn about audit and consulting retail services? Our consultants will be happy to help you. Fill out the form, and we will call you back.
What does store clustering give a retailer?
Cluster analysis suggests in which groups and which narrow subcategories should be developed. That is, you do not need to try to insert the same assortment at all points. And it should be understood that a certain group of stores is visited by its own category of customers. And they are looking for products with certain properties. For example, a diaper of a low price category, milk of a certain fat content, products in small packages, or other narrow properties. And clustering gives the answer that it is in this group of stores that it is worth developing this category of goods and in no others.
In a large chain, it is difficult to operate with the properties of goods and targeted assortment at the level of one retail outlet. Since it will be resource and time-consuming. But by choosing groups of stores that are similar to each other in the structure of consumption of certain goods, we can form a similar assortment. Thus, we get a specialized offer for buyers with similar needs.
The clustered approach to targeting assortment optimizes the number of properties to manage. In addition, clustering gives an understanding of what is so special about these objects and what attracts buyers to these outlets.
C4R solutions for clustering in retail
Cluster analysis is already built into C4R's category management solutions. We help and conduct research to identify patterns in customer behavior, identify significant attributes in each of the product categories. Our team has over 15 years of experience in consulting retail business processes.
C4R solutions help you build a better assortment and increase your sales. For example, in the joint case of the Magnum Cash & Carry retail chain and C4R, an 8% increase in turnover was achieved thanks to the introduction of an assortment management system.
If you want to learn more about clustering, features of transferable demand, solutions for automating retail, logistics and distribution processes, as well as audit services for retail and consulting, write to [email protected] or fill out the feedback form. Our experts will contact you.
The author of the article is Andrey Shevchuk,
Head of Business Consulting C4R,