It is essential to identify the segments to which your customers belong, in the long term it is even more relevant to analyze the transition matrix in order to improve your marketing strategy. The challenges of segmentation: setting the foundations  If we compare data exploitation to building a house, then customer segmentation will allow us to set the foundations. Firstly, segmentation will allow you to understand the structure of the customer base and to identify the different groups to be animated. The resulting indicators will allow you to build a strategy based on customer value and loyalty. However, simply looking at the picture of the segmentation at a given time may not be enough. You must also pay attention to the transition matrix. The interest of the transition matrix Indeed, when analyzing segmentation, it is important to take into account long-term dynamics. In particular, it is essential to study the flow of customers between two segmentation periods, which is called the transition matrix, or flow matrix. The idea of the transition matrix is to determine in what proportion the customers of a segment at a given moment in year N will migrate to another segment in N+1. Two types of matrices can be calculated. The volume matrix allows us to know how many customers have moved from segment A to segment B between two periods. The percentage matrix will allow to measure the proportion of customers migrating from one segment to another. Example and analysis of a datacadabra transition matrix As an example, this percentage transition matrix will allow us to see the proportion of customers who move from one segment to another between the segmentation calculated in N-1 and the segmentation calculated in N. We are thus interested in three types of movements. The first key piece of information to analyze is the stability of clients, i.e. the proportion of clients who remain in the same segment between two dates. This information will allow us to validate the level of loyalty of high segments (VIP / TBC here). We can then analyze the upward flows, i.e. the customers who have seen their segmentation level increase over time. This will allow us to measure the extent to which the actions taken by the brand have improved the quality of the base over time. Finally, the downstream flows, i.e. the customers who move from one segment to a lower one, will allow us to see to what extent attrition is an issue to work on as a priority. If we look in detail at the results presented in this matrix, we can see that 43.6% of VIPs have remained VIPs, 34.3% of very good customers (TBCs) have remained TBCs, etc. These levels of stability are fairly average and highlight a problem of loyalty among the core target. It will be necessary to implement specific actions on these segments to increase their commitment. It should also be noted that 53% of new customers (N) become inactive after one year. The nursing plan could therefore be improved in order to improve the quality of recruitment over time. We also note that nearly 25% of good customers (BCs) and more than 50% of COs become inactive after one year. It would therefore be interesting to work on the anti-attrition processes for these targets. To sum up, the analysis of the transition matrix will make it possible to identify the major issues on which to focus the animation strategy (nursing, anti-attrition, reactivation, loyalty of the core target group, etc.) and to determine the additional analyses to be carried out to optimize the action plans. How does datacadabra support you on the subject?  Within the Segment module, the different segmentation methods natively propose the transition matrices over different periods. They are enhanced with automated comments to measure the observed performance. Want to know more? Do not hesitate to contact us or to ask for a datacadabra demo.

When we talk about marketing targeting, the first driver we think of is segmentation. Given the current trends, these two subjects will most certainly be at the forefront of the news in 2022. How can segmentation help you in your targeting? If we compare data exploitation to building a house, then customer segmentation will allow us to lay the foundations. Firstly, segmentation will enable us to understand the structure of the customer base and to identify the different groups to be animated. The resulting indicators will allow you to build a strategy based on customer value and loyalty. However, relying solely on the photo of the segmentation at a given moment can sometimes prove insufficient, which is why targeting is necessary. Once the segmentation is obtained, how can it be activated with targeting? This is where the different monitoring tables resulting from a segmentation will help you in defining your animation strategy. First of all, identifying the value of each customer segment will enable you to define the budgets per segment in terms of both commercial investment and generosity. This first step will necessarily condition all the actions that will be defined thereafter. Next, the transition matrix will make it possible to determine the major strategic issues per segment. This will make it possible to identify, for each segment, whether targeting should focus on loyalty, attrition, increasing the frequency of visits, increasing the average basket, reactivation, etc. The last element of customer knowledge that will improve the quality of targeting is the characterisation of the groups obtained. As shown in our article “Characterize your segments and adapt your strategy“, characterisation will make it possible to define numerous complementary customization and targeting drivers. This will therefore enable a more customized response to customers’ expectations and improve performance. We have seen our clients achieve performance gains around 5 to 10% of their turnover thanks to the use of data.  You want to improve your targeting but you don’t have the right tools? With datacadabra, you can build your customer segmentation in just a few clicks using the Segment module. Our different segmentation methods allow you to identify the similarities and differences of your segments in order to customize your marketing campaigns.  Once your segmentation has been implemented, the Target module allows you to exploit all the richness of your data (raw data, segmentations, scores) in order to define your most relevant targeting criteria. With one objective: send the right message to the right customers. Want to know more? Do not hesitate to contact us or to ask for a datacadabra demo.

When developing your marketing strategy, it is essential to define your different segments in order to build a solid analytical base.  When you want to optimise your marketing and CRM strategy, you often start your customer knowledge work by implementing your analytical base. The first step is to characterize your segments by setting up a customer segmentation. This will enable you to identify the main groups to be managed and to define the main actions to be carried out on each of them. In order to better understand the characteristics of each group, it is often necessary to characterize your segments. Indeed, whatever the segmentation carried out, it is interesting to understand the profiles of the different segments in order to improve the customization of the segmented animation plan. Example 1: Differences in consumption per channel  In terms of implementation, characterizing segments is based on an analysis of customer profiles and their consumption according to the segment to which they belong. It will thus be possible, for example, to analyse the distribution of segments by consumption channel. The table above shows that Very Important Customers and Very Good Customers are over-represented among mixed customers. While New Customers are over-represented among exclusive web consumers and Occasional Customers are more likely to buy in-store. This information will allow either to direct customer communications towards the preferred channels of each segment, or to favor omnichannel by proposing offers in favor of the complementary channel. Example 2: Differences in product consumption In the same way, it will also be possible to analyse the consumption of the different segments by product family. The table above shows that Very Important Customers are over-represented in the Accessories family and especially in the Apparel family. The Very Good customers are over-represented in Accessories. We will therefore personalize the product offer according to the segments. In parallel, we can deduce that diversification in terms of products is also a vector of loyalty. It will therefore be appropriate to highlight certain products to the soft core segments to increase their knowledge of the brand, their consumption and therefore their loyalty. Example 3: Differences in sociodemographic profiles Another element that will be important in understanding the different segments is the analysis of their sociodemographic profile. This will enable us to understand the differences between the different segments in terms of age, sex, socioprofessional category, standard of living, etc. The graph above gives an example of the characterization of loyal segments versus the population through GeoTypo. We can see here that the active segments are over-represented in rather urban and SPC- areas, whereas they will be under-represented in SPC+ districts. This socio-demographic information will also make it possible to improve the digital acquisition process by focusing on the characteristics of the most loyal segments. How does datacadabra help you characterize your segments? As we can see, the characterization of segments will make it possible to find numerous drivers for improving communication. Within datacadabra, the Describe module will allow you to work on different types of profiles, on the brand’s own data or on Open Data.  Want to know more? Do not hesitate to contact us or to ask for a datacadabra demo.

Suspect, prospect, customer, it is now essential to master the customer life cycle or sales funnel to make your sales strategy profitable.  What is the customer life cycle?  The customer life cycle can be defined in different ways in marketing. Either a literal definition corresponding to the events that the customer will experience in his life (marriage, birth of a child, etc.), or a definition relating to the relationship between the customer and the brand. In this second case, which is of interest to us here, the customer life cycle will designate the different stages in the evolution of the relationship between the customer and the brand.  A relationship that starts long before the purchase: the suspect and the prospect The relationship between an individual and a brand starts long before the first transaction. Even before the first interaction between the contact and the brand, if the individual corresponds to the target established by the brand, he will be considered a suspect. Subsequently, as soon as the first exchanges start and the brand is able to identify the individual, he will have the status of a prospect. The brand’s first challenge will then be to succeed in optimizing the conquest phase to transform this prospect into a customer. These first three phases already require numerous actions on the part of the brand to increase the performance of its recruitment processes. In particular, test and learn phases and the creation of recruitment scores will make it possible to improve the relevance of the first stages of the customer life cycle. The first transaction: the customer  Because once the contact has carried out his first transaction and acquired the status of customer, he will then be able to follow a path that will take him from new customer to loyal customer to departing customer to inactive customer and finally to reactivated customer. When you build your customer segmentation, and you study the segmentation flows over time, you can identify the main paths followed by customers.  A direct impact on business investment  The interest of knowing the customer life cycle well and being able to measure and optimize the commercial investments to be made at each stage. Obviously, from one sector to another, the customer life cycle lasts more or less time. For example, on the Internet, customers are quite volatile, whereas in the insurance sector customers are much more loyal. The role of data science in the customer life cycle  At the various key moments in the life cycle, we will be able to use data science techniques to help the business in its actions. For example, for acquisition, we can work on recruitment scores or identify the priority profile to recruit. In order to correctly develop its customer file, it will be possible to segment it. We will set up targeted actions on the core target to increase loyalty. Another important point: we will anticipate attrition through dedicated scores. Finally, we can prioritize certain targets to be reactivated by using models specific to this problem. Data Science will therefore be able to interfere at different moments in the customer life cycle, either through descriptive analyses or predictive analyses. The most important thing is to prioritize the actions to be taken so that the analysis system is as effective as possible. In the end, it’s a bit like building a house. Within datacadabra, the Describe and Predict modules will allow you to work on these issues in order to optimize your action plans. Want to know more? Do not hesitate to contact us or to ask for a demo of datacadabra.