Customers have different needs and behaviours. By grouping together those that are similar, you can adapt your approach to suit those differences. You can do this by targeting the customers who provide maximum value, encouraging new habits in those who cost you and tailoring your service to the different ways people use it.
Understanding the needs and behaviours of thousands or millions of customers is a huge undertaking, let alone knowing what the most effective way to group them is. Even if you can do this, you still have to identify and take advantage of the opportunities within your segmented data.
Peak’s data machine analyses your customer information including transactions, spending, demographics and decision-making. It uses a variety of clustering techniques to spot key customer groups and creates predictive models to work out how different groups will behave in different scenarios.
The Techie Bit
We use supervised clustering algorithms based on groups of classification trees to segment customers by different marketable traits. These algorithms look at which variables are the best predictors of lifetime value and use those to build a powerful predictive model. To improve this segmentation, we can augment your data with our own external data sets. When demographic data is scarce, we can segment your customers based on the likelihood of individuals reaching a certain high-performance metric or by recent performance, and present the key behavioural differences to you.
Customer segments, along with their needs and behaviours, are presented to you in a dashboard. Our data machine also provides insights and recommended courses of action for you to take with each group of customers, while our analysts provide additional, plain-English support and work with you to put actions into place. You can then use the dashboard to track the change in performance of each customer group over time.