Most consumer goods companies stop earning millions of dollars per year by continuing to implement the same shopper activation solution in all the chain stores that sell their products. The first step? Segment your key account stores based on the mission and purchasing habits of their customers.
Are all stores of the same brand or flag (account) selling exactly the same?
No! Although the account does not make major differentiations and some stores seem the same from the point of view of image and size, they will surely have big or small differences: they will be in different regions and cities, different areas within the same city and, above all, they will have shopper profiles. with different purchasing missions even for the same category. It should also be noted that the performances of a category and its respective brands change from one location to another. In summary, these analyzed elements affect product demands and purchasing behaviors very differently from one store to another.
This being so, it is essential to look for ways to group (segment) stores that present similar behaviors, especially associated with profiles and purchasing missions, which will allow us to define differentiated assortment, space, communication, and promotion strategies.
We could argue that the ultimate goal is to create a unique strategy for each store and technology already allows us to do so.
But why key account stores are not segmented correctly?
The most frequent answers to this question are:
We haven't thought about this.
We don't know how to do it.
The account will not give us sales data per store.
The account has its segmentation model
I think we already covered the first answer with the introduction of this article. Let's now look at the other three.
How to segment the account stores?
There are several ways to segment, from the most traditional using demographic and geographic statistics to the most current ones, which are what we are using at TMC Consultores, where we incorporate Machine Learning (ML) algorithms that can process high amounts of data, generating much more precise and dynamic results (which are continuously adjusted) and, as a result, significantly increase sales volumes.
For obvious reasons, we are going to focus on these new technologies in the rest of the article.
Store segmentation with ML algorithms
The data provided by each buyer when visiting a store has tremendous value when processed with the rest of the purchase tickets that are generated over a period of time. 12 months is recommended. Algorithms can detect purchasing patterns from the millions of data generated and group stores that have similar behaviors into clusters or segments. Artificial Intelligence processes convert data into knowledge.
Store Segments
Once the stores were grouped and analyzed, we detected interesting patterns that we could hardly obtain through more traditional means. We are going to see segments of stores that not only have similar sales of references (SKUs) but also similar purchasing behaviors. For example, we will observe by segment:
References that sell more on certain days of the week or times of the day than others.
Correlations of when a category, brand, or reference is purchased, with other categories, brands, or references.
Different purchasing missions for the same category. For example: in one store we can observe predominantly planned purchasing and in another store, impulse purchasing.
If we additionally cross-reference this data with the sales volume of each store, we will be able to adapt and focus differentiated actions for groups of similar stores, and include and exclude stores, according to the strategic importance that each one may have.
The Success Photo by Segment
Perhaps the most obvious and immediate thing to implement is the design and execution of the Success Photo by store segment, for the same account. That is, define the assortment, brands and main references, planograms, displays, and even prices, when advisable, by segment.
The data generated by the Retailer can be continuously incorporated into the algorithm so that it generates adjustments to references and spaces and even moves a store from one segment to another when it detects changes in purchasing behaviors. Once again, the keyword is knowledge.
But the Retailer has its segmentation model
Of course, but it is not necessarily built based on your strategic objectives or thinking about your categories and brands.
If you consider that the Retailer's segmentation adapts to your needs and requirements, great. Just take advantage of it. But if not, better make your segmentation.
Execution Capacity
It requires greater effort and more resources to execute actions in a differentiated manner in 4 store segments than in just one. However, all the companies with which we have worked on this issue corroborate that it is an investment with a very rapid recovery.
The results in incremental sales and investment optimization of Trade Marketing generate a very high return.
Conclusions
The commercial functions of Trade Marketing and Key Accounts today have ML technologies at their fingertips, which will allow them to take advantage of existing information like never before to segment the accounts' stores, to focus initiatives and strategies of brands where the target Shopper and the predominant purchasing mission are.
If you want to know our store segmentation ML algorithm and increase your sales by making the necessary adjustments, write to me at jmdominguez@tmcconsultores.com, and let's schedule a meeting.
Written by Juan Manuel Domínguez R. CEO of TMC Commercial Consultants. If you are interested in learning about TMC's consulting or training products in this matter, write to us at contacto@tmcconsultores.com and we will immediately contact you.
If we are not yet connected on LinkedIn, it will be a pleasure to have you in my network of contacts https://es.linkedin.com/in/juanmanueldominguezr