Model retention strategy
Posted: Sat Dec 21, 2024 5:24 am
Social selling: finding your target audience
In order to meet and 'connect' with your target group, it is important that you know how to find them. You know your target group and you know (if all goes well) who the decision-making units are and on which social media they can be found. Make smart use of the search engines and filter options to really reach your specific target group, so that your message can have the maximum effect.
The book 'Social selling in 60 minutes' (aff.) gives some life-pro-tips for specific searching on LinkedIn that I of course do not want to withhold from you. These techniques fall under the heading of boolean search (a well-known term for the programmers among us; true?).
The logical start is to look up the company you want to do business with and then look more specifically for the employees who could be of interest to you. But sometimes you don't know yet which specific company you want to enter into a relationship with. Then the following 'tricks' may offer help:
LinkedIn sees a space in keywords like 'and' and therefore filters out part of your target group. By putting your search combination between double quotation marks ("interim project manager"), LinkedIn searches for the exact word combination.
Sometimes you want to exclude keywords, you do this by putting NOT in front of the desired excluded word, such as “commercial director” NOT interim.
Multiple options in your search are also possible, you do this with OR (project manager OR contract manager)
Do you really want to search specifically? You can combine the above options using brackets, such as: "commercial director" (transport OR logistics) AND (hbo OR mbo) NOT interim
Who is this book for?
Social selling in 60 minutes is written sympathetically and easily readable, not too formal, with nice personal input from the author. However: the book (especially the first half) is mainly aimed at beginners in marketing. For example, a large part of the (small) book is about personas, objectives, proposition and explanation of the social media channels. For the starters among us a recommendation as a holiday book, perhaps the professionals would be better off going for the more in-depth version that colleague Judith Eversdijk wrote about earlier : Social Selling Masterclass by Carola Rodrigues.
As a company, you have incredibly valuable information at your disposal if you are able to predict whether customers will cancel or leave. Especially if you can also predict this at an individual level. Even better would be if you could also discover which customers are valuable to your organization and in whom you should therefore invest time and energy. You would also want to know which resources you should use to retain the most important customers. So that you can use these insights to execute marketing campaigns much more effectively. To make the picture completely perfect, you want to know how you can make a nice offer to the customers who ultimately decide to leave in order to win them back. In short: what does the ultimate data-driven retention strategy look like? In this article I will explain which predictive models you should use to achieve this.
4 predictive models
So how do you arrive at that ultimate retention strategy? In this article I will explain which four different predictive models you need to tackle to arrive at a data-driven retention strategy. When applying the retention model correctly, you achieve optimal customer retention, whereby you steer on value. This strategy is applicable to every sector and is interesting for everyone who works with data: from marketers and data scientists to sales managers and strategists.
In the image below you can see the different steps you need to take to arrive at the ultimate strategy. We start by predicting which customers will leave (churn), then predicting the value of the customer (CLV). Then we predict what we need to do to retain customers (NBO/NBA), finally we predict who we need to approach from the laos telegram data customers who have left anyway to retain them ( saves ). With these predictive models we can set up a data-driven retention campaign!
1. Churn modelling
What is churn modeling?
Let's start at the beginning. What exactly is churn modeling? Churn is a term that indicates which people (customers) change service or supplier in a certain period. With churn modeling you can predict how likely it is that a customer will switch or cancel. Churn modeling does not only apply to companies with subscription forms as a product or service. Churn can also be defined in retail, for example: churn is not buying anything in a period of three months.
Training the algorithm
The first step to the ultimate retention model is predicting what a customer is planning to do. Often you first think about the risk that the customer will switch to the competitor. Or the chance that the customer no longer wants to use the service or product. But to be able to steer really effectively, you want to go a bit further. You not only want to calculate the chance of churn, but also calculate what the customers with a low chance of churn are going to do. We can distinguish between three types of customers.
In order to meet and 'connect' with your target group, it is important that you know how to find them. You know your target group and you know (if all goes well) who the decision-making units are and on which social media they can be found. Make smart use of the search engines and filter options to really reach your specific target group, so that your message can have the maximum effect.
The book 'Social selling in 60 minutes' (aff.) gives some life-pro-tips for specific searching on LinkedIn that I of course do not want to withhold from you. These techniques fall under the heading of boolean search (a well-known term for the programmers among us; true?).
The logical start is to look up the company you want to do business with and then look more specifically for the employees who could be of interest to you. But sometimes you don't know yet which specific company you want to enter into a relationship with. Then the following 'tricks' may offer help:
LinkedIn sees a space in keywords like 'and' and therefore filters out part of your target group. By putting your search combination between double quotation marks ("interim project manager"), LinkedIn searches for the exact word combination.
Sometimes you want to exclude keywords, you do this by putting NOT in front of the desired excluded word, such as “commercial director” NOT interim.
Multiple options in your search are also possible, you do this with OR (project manager OR contract manager)
Do you really want to search specifically? You can combine the above options using brackets, such as: "commercial director" (transport OR logistics) AND (hbo OR mbo) NOT interim
Who is this book for?
Social selling in 60 minutes is written sympathetically and easily readable, not too formal, with nice personal input from the author. However: the book (especially the first half) is mainly aimed at beginners in marketing. For example, a large part of the (small) book is about personas, objectives, proposition and explanation of the social media channels. For the starters among us a recommendation as a holiday book, perhaps the professionals would be better off going for the more in-depth version that colleague Judith Eversdijk wrote about earlier : Social Selling Masterclass by Carola Rodrigues.
As a company, you have incredibly valuable information at your disposal if you are able to predict whether customers will cancel or leave. Especially if you can also predict this at an individual level. Even better would be if you could also discover which customers are valuable to your organization and in whom you should therefore invest time and energy. You would also want to know which resources you should use to retain the most important customers. So that you can use these insights to execute marketing campaigns much more effectively. To make the picture completely perfect, you want to know how you can make a nice offer to the customers who ultimately decide to leave in order to win them back. In short: what does the ultimate data-driven retention strategy look like? In this article I will explain which predictive models you should use to achieve this.
4 predictive models
So how do you arrive at that ultimate retention strategy? In this article I will explain which four different predictive models you need to tackle to arrive at a data-driven retention strategy. When applying the retention model correctly, you achieve optimal customer retention, whereby you steer on value. This strategy is applicable to every sector and is interesting for everyone who works with data: from marketers and data scientists to sales managers and strategists.
In the image below you can see the different steps you need to take to arrive at the ultimate strategy. We start by predicting which customers will leave (churn), then predicting the value of the customer (CLV). Then we predict what we need to do to retain customers (NBO/NBA), finally we predict who we need to approach from the laos telegram data customers who have left anyway to retain them ( saves ). With these predictive models we can set up a data-driven retention campaign!
1. Churn modelling
What is churn modeling?
Let's start at the beginning. What exactly is churn modeling? Churn is a term that indicates which people (customers) change service or supplier in a certain period. With churn modeling you can predict how likely it is that a customer will switch or cancel. Churn modeling does not only apply to companies with subscription forms as a product or service. Churn can also be defined in retail, for example: churn is not buying anything in a period of three months.
Training the algorithm
The first step to the ultimate retention model is predicting what a customer is planning to do. Often you first think about the risk that the customer will switch to the competitor. Or the chance that the customer no longer wants to use the service or product. But to be able to steer really effectively, you want to go a bit further. You not only want to calculate the chance of churn, but also calculate what the customers with a low chance of churn are going to do. We can distinguish between three types of customers.