In the first part of this blog, we highlighted the importance of data utilisation to the success of data-driven marketing initiatives and introduced the Avaus data classification model. Here, we will look at practical ways to use this tool in realising the value locked up in your companies’ data assets.
Let’s take the example of a B2C retail company selling consumer goods that have three typical business objectives at different operational levels: passive customer retention, loyal customer engagement, and a narrower purpose regarding the effectiveness of email marketing.
Win-back – how to prevent customer loss
It is self-evident that keeping a current customer is more profitable than acquiring a new one. So how can we identify the risk of losing a customer from customer data? Without doing anything in the way of predictive analytics, it’s easy to set business rules defining what a lost customer is and enrich data by creating customer segments based upon these.
This segment data provides insights over the longer run and enables the creation of actionable target groups for win-back campaigns. Applying the data categorisation described in my earlier post, we can utilise transactional data to identify the risk.
The high-precision approach is to use advanced analytics techniques and calculate the individual churn probability for each customer, rather than rely on the broad-brush impression given by traditional rule-based segmentation.
The crucial thing to identify is changes in customer behaviour that indicate that they are growing passive. For example, in a very simple segmentation model, the following rule can be set: if a customer’s purchasing frequency has fallen 50% in the previous quarter, the business rule definition shifts the customer to the high-risk segment.
Before you execute a set of retention campaigns for this segment, it is split into target and control groups to enable campaign efficiency analysis later on. The campaign results may look like this:
The campaign analysis table above shows the percentage of customers in each segment who remained and how many moved on compared to Q1. In this case, the retention campaigns were successful, because 25% of the target group customers in the high-risk segment shifted to medium and low risk from Q1 to Q2, while there was a 20% loss of those in the control group. The role of data enrichment and analytics was fundamental, as it helped to identify the customers with high churn risk.
Engagement tactics for loyal customers
The merging of receipt and customer basic data has been key to gaining customer insight and running targeted marketing actions for decades. Campaigns using lots of airport advertising are known to generate loyal clients as the demographic there is usually successful.This merging reveals individual purchasing preferences and behaviours. It’s possible to identify patterns for specific items and their average frequencies. These may be determined as follows:
This kind of purchasing behaviour applies to a vast number of product groups from cosmetics and detergents to light bulbs and ventilation system filters. These are any non-grocery products purchased in a cyclical pattern that doesn’t appear on shopping lists until used up. When you’re likely to be running out of coffee filters, a smart retailer could make an approach with a related offer.
Accuracy in the identification of typical purchasing patterns can be enhanced using digital footprint data from a web analytics or DMP tool, which will often reveal the visitor’s purchasing intentions for the near future.
Based on personal experience, as a loyal customer of all the major retailers in Finland, the customer engagement and brand behaviour implications of all this have yet to be fully realised. What is holding marketers back from being more creative and savvy with data?
The creation of a set of such customer engagement marketing programs would not merely differentiate the retailer’s personalised marketing from those of its competitors. It would bring its customers enormous gratification by offering them content precisely tailored for them.
In the next post of this blog, we’ll continue with the tactics of email marketing optimisation.