Artificial intelligence for marketing
Acquiring new customers is more expensive than retaining your existing customers. This is common knowledge for each CMO. However, most marketing teams cannot quantify the value the company will get during the relationship with a customer, so they have to guess how much money they should and can spend on customer experience to retain this customer. Furthermore, most marketing teams do not know which of their measures are effective for improving customer retention or if there are different groups of customers in their customer base that should be treated with different retention measures. In this article you will learn how to use artificial intelligence for customer retention.
Questions that I will answer in this post:
- How much will a specific customer or a group of customers be worth throughout their entire customer lifetime?
- How can I increase the customer lifetime value of our customers?
- Which of my measures to increase customer retention are most effective? Which are inefficient?
- Which retention measures should I use for which customer groups?
- Which customers are more likely to be loyal?
- How do I stimulate the use of my digital product or the digital service I offer?
How much will a specific customer or a group of customers be worth throughout their entire customer lifetime?
We build machine learning algorithms that predict the future customer lifetime value of your customers on an individual level. This means you will know how much money you will earn with each customer until the end of your relationship. Your marketing team will be able to seperate high value customers from mid or low value customers and therefore be able to treat those groups differently. As we push this actionable information right into your CRM, your marketing team can build automated campaigns for customers with different values.
How can I increase the customer lifetime value of our customers?
Our algorithms will provide the information which of your product or service combinations, customer journeys and experiences or customer behaviors are resulting in high lifetime values. Equipped with this data you can then optimize your offering, your segmentation and your campaign strategy. Your marketing team will be able to deliver important customer experiences to more of your customers and try to influence customer behavior in a way that those customers get more value from your offering. If you increase the value customers get from your customer experience, your customers will stay longer and spend more.
Use Case: Artificial intelligence for customer retention
A day in the life (before)
Scene or situation:
Michael A is the CMO of a multinational media company offering various journalistic products to their subscribers. As they offer an immense amount of different news subscriptions, magazines and journals on varied topics, each customer’s consumption pattern is different. Michael wants to increase the average time a customer stays with his company and thereby improve the average customer lifetime value. To retain his company’s customers longer, he creates a customer lifecycle team consisting of marketing experts and gives them the task to improve retention rates and the average retention period of the customer base.
Desired outcome:
Create new experiences that positively influence those variables that are the strongest drivers for customer retention. Furthermore, find those of the already implemented experiences that make the customers stay longer and therefore spend more with the company. Make sure that more customers experience them.
Attempted approach:
The new customer lifecycle marketing team analyzes the experiences a customer will have during their customer lifetime. After plotting these experiences on an experience canvas the team brainstorms for ideas on how to make the customers stay longer. During their discussion some main ideas crystallize: Explain the features of the products better to the customers so they can get more value from their usage, increase the customer engagement with the content platform to make them read more and make them use more than one journalistic product. The team decides to focus on two tasks in the next quarter. First, they want to improve the onboarding journey of new customers by adding more emails to the already existing onboarding journey. In these emails they want to explain the different features of the user interface of the website and their app and teach the customers about the various formats that can be read with their subscription. The second experience they decide to build is to send regular emails informing customers about products they are not using yet, such as other magazines or journals. The team starts to write briefings for the creative department and prepares their marketing automation tool. After a month the copy and the creatives for the emails are delivered by the creative department. The marketing team implements the new onboarding emails into the existing onboarding journey and sets up a new journey for suggesting not yet known products to their customers.
Interfering factors:
The team had some clever ideas that maybe could help to increase customer retention. But as the team has no guidance from a machine learning algorithm, they are basically guessing what could be a driver of customer retention. As retention can only be measured over time from cohort analysis, it will take months to figure out if the experiences they built are successful in extending the average retention time of their customers.
Economic consequences:
As it takes a long waiting time until retention measures can be evaluated (generally a few months of monitoring of the cohorts), the team will only know whether what they did made sense after months of investing their efforts. This results in high costs for the company to evaluate which of those initiatives work or not. Furthermore, surely there are other variables that would be easier to manipulate to optimize retention. Besides wasting budget on guessing efficient retention measures, there are high opportunity costs associated with wasting months of not optimizing retention experiences. Each customer who churns needs to be compensated for by customer acquisition, which is way more expensive than retaining the customer. Retaining a customer longer not only increases the CLV of that customer as the subscription period is prolonged, but also enables more opportunities for cross- and upselling as well as more retention measures to keep the customer even longer.
A day in the life (after)
New approach:
Michael decides to partner with aaimo and orders a machine learning model to better understand customer retention. From the output of aaimo’s machine learning algorithm, Michael’s team has learned that there is a strong relationship between the daily consumption of the services and the retention time. The lifecycle marketing team conceptualizes a new daily newsletter that suggests the most viewed articles to each customer, thereby increasing the average daily consumption of their service and eventually the retention time of the customer base. Furthermore, they learned that the retention time drops rapidly if a user has interacted with their product less than five times a month. As they know that the monthly usage should not drop below five times, they create a second customer journey for all customers that have less than five uses per month. Those customers get a series of communications that show recommended articles, podcasts and interviews based on past usage behavior and individual customer’s interests. A third interesting insight from aaimo’s machine learning model is that the existing onboarding journey has no positive effect on the retention rate for subscribers of magazines. Michael’s team therefore decides to build a new onboarding journey for subscribers of magazines specifically customized for this product.
Enabling factors:
Michael’s lifecycle marketing team now has a deep understanding of what factors are influencing customer retention. When conceptualizing new experiences they can use this knowledge to focus on the most important factors. Furthermore, the team now not only knows which of the already existing customer experiences work, but which type of experience works best for which customer type . They can further individualize the experience of their customers by creating more customized experiences or tweaking existing ones for distinct customer groups.
Economic rewards:
A few months after the implementation of the new experiences Michael receives a report from the analytics team. The cohort analysis of the retention rate shows that the new experiences positively influence the retention rate of their customers. Michael is satisfied with his customer lifecycle marketing team and tasks them to build more experiences based on the insight of aaimo’s machine learning model to further improve customer retention for their company. As their customers’ average retention time increased, so did their customer lifetime value and the company’s profit.
Which of my measures to increase customer retention are most effective? Which are inefficient?
Most companies have a battery of customer journeys for retention that each customer experiences during their customer lifetime. Even if those customer journeys have been thoroughly tested and evaluated, most companies can only prove that the battery of customer journeys as a whole is improving the retention rate. We can create a machine learning algorithm that will not only tell you which of those individual experiences are effectively improving the retention rate, but how strong the effect of each customer journey is. You can then improve on the experiences that matter and replace those that don’t.
Which retention measures should I use for which customers?
All of your customers are individuals. They have different needs and preferences. Your retention strategy should reflect this. For some of your customers a specific customer journey will improve their retention rate, for others this experience might have no effect or worse. If you send a communication with the goal of improving the retention time, that has no effect, you are wasting a chance to deliver a relevant experience. The artificial intelligence we build for your marketing team will provide you with insights on which customers should be targeted with what retention treatment. This information will make it possible for your marketing team to trigger automated retention journeys from your CRM to deliver the right retention measure to the right customers.
Which customers are more likely to be loyal?
We can build an artificial intelligence that predicts which of your customers are more likely to be loyal. Your marketing team can then design a different loyalty program or a specific campaign for these customers. This will ensure that you focus your retention budget on those customers that are the most influenceable by your retention measures.
How do I stimulate the use of my digital product or the service I offer?
If you offer a product or a service whose value lies in regular use or even increases when used more often, knowing which customers drop below a certain usage threshold is very useful. This information can be an early indicator for churn. If customers with decreasing usage are labeled in your CRM, your marketing team can build experiences to re-engage them. Finally, your customers can experience the value of your service or product again and stop thinking about churning.
Many companies do not have the resources or the expertise to make optimal use of their data for marketing operations. We help companies to understand what AI use cases are most important for their specific situation, industry and business goals. We analyze if the data of your company is sufficient for AI use cases and run machine learning models as a full service. We deliver the output right into your CRM where your marketing team can use it to build and automate customer experiences.