Propensity scoring predicts the likelihood of a particular type of customer behaviour, from how they will respond to specific messages to whether they are in market for a particular product. Knowing how, when and why a customer is going to act in a certain way lets you deliver the right message at the right time, enabling you to tailor your content to the specific needs of a business or customer – ultimately improving your advertising and marketing efficiency.
Which customers will be most interested in your next promotion? What customers are at risk of churn? What leads should you focus your resources on? What channel should they be contacted on, and when will they be in market? Knowing which point in the customer journey particular content has the most impact requires a full understanding of a customer’s purchase behaviour, often requiring the unification of multiple datasets.
The Peak AI System contains a multitude of machine learning techniques at its disposal, utilising those that are specific to the scoring metric required. For example, we use supervised learning approaches to predict if customers are likely to purchase unique product categories, or more advanced deep learning techniques to score a customer against an entire product catalogue. Traditional techniques, such as linear or logistic regression, also enable us to identify the right time of day or week to contact a customer.
Our AI System provides scores for each customer, for any particular outcome that is important to your business. This enables you to filter or segment your customers as best suits your business needs. For example, you can target the optimal customers for a particular product or event with highly personalised content, send special offers to customers who are at risk of churn, or prioritise your resources (such as sales staff) towards leads who are most likely to convert.