In our last installment of the Big Data & Customer Intelligence series, we ended up with a discussion of customer segmentation by using predictive analytics to lay focus on the customer groups most open to potential influence. In this article, we’ll outline how you might go about developing a customer influence strategy. This is an important element of Computational Marketing.
The overriding goal in utilizing customer segmentation is to use data to get closer to customers – this ability is every marketer’s dream. By getting close to customers, you gain the opportunity to deliver services and marketing with unprecedented precision and accuracy, thus meeting or exceeding customer expectations. A happy customer is a valuable customer. In using predictive technology, your company can maximize customer lifetime value and retention. Customer retention is key here, it costs 6-7 times more to attract new customers than it does to retain existing ones. Fortunately, a number of new big data start-ups have entered the realm of providing technology to influence customer behavior. A good example of a new vendor in this space is Retention Science.
So just how does the science work? Here are some primary points to remember in utilizing big data to achieve proactive customer influence:
- Analyze and predict customer behavior.
- Personalize customer retention campaigns.
- Predictive algorithms determine a customer’s purchase probability for each available offer and deliver the lowest discount needed to convert.
- Provide custom product recommendations through analytics to determine what product best satisfies the customer (recommender systems).
- Use statistical models to provide a holistic view of customer activity and interaction with your company.
Once you’ve instituted a comprehensive customer influence strategy, you can anticipate a number of concrete benefits such as increased spending from existing customers, improved business margins, and increased campaign conversions.
[The next in the series will discuss determining customer value]