A customer's likelihood to still be your customer is based on their patterns of engagement with your brand. By identifying customers who are breaking their normal patterns, you can intervene and save the relationship at the most opportune time. Zaius' customer winback model shows the relationship between days since engagement and likelihood to still be your customer, so that you can find fading customers before they churn.
Zaius' winback model is optimized for brand engagement, meaning that both new shoppers and previous purchasers are eligible for evaluation.
Using the Churn Prevention Explorer
Customer Engagement Trend
The cone plot represents the relationship between time since last engagement and likelihood to still be your customer for your average customer. Each customer takes a different path through this cone based on their own patterns of engagement:
- A frequently-engaged customer’s liveness probability drops more quickly as their engagement slows down (the lower border of the cone).
- A less-engaged customer’s probability drops more slowly, as they already had a longer time between engagements (the upper border of the cone).
The cone shows your middle 60% of customers. 20% of customers may fade more quickly than shown (have a steeper drop), while 20% may fade less quickly than shown.
The cone also includes areas denoted as Engaged, Winback, and Churned. These inflection points are specific to your brand's data.
- The Winback area begins where an increase in a customer's likelihood to still be a customer yields a correspondingly larger increase in their likelihood to re-engage in the next year—this is the point at which you want to start a winback campaign.
- The Winback area ends where an increase in a customer's likelihood to still be a customer yields a correspondingly smaller increase in their likelihood to re-engage in the next year—at this point, the customer can be considered churned.
Distribution of Customer Likelihood
The histogram at the bottom of the explorer shows the distribution of your customers by their likelihood to still be a customer. The default view is your full known customer base (it does not include anonymous shoppers). Build new segments or evaluate existing segments by clicking on the "+ Segment" button to apply them to the histogram.
The histogram helps you understand the composition of different audiences. For example, if your newsletter audience skews Churned, it could impact your open rates and deliverability. If you have high-value customers slipping into Winback, you may want to approach your relationship with them uniquely.
Scored vs Unscored Customers
A box on the right of each analysis indicates the number of unscored customers. An unscored customer does not have a value for the customer zone (Engaged, Winback, and Churned). A customer is unscored if they have only had activity on a single day and they have never made a purchase. This percentage will vary if you apply different segments to the Distribution of Customer Likelihood chart.
Just like it can be important to monitor the percentage churned in your key targets, you should also look at the percentage unscored, as these customer may not be a good fit for your brand and can likewise impact open rates and deliverability in campaigns, or consume marketing resources in other forms of targeting (like segment sync). If a customer has been known to you for a while but is still unscored (has no value for the Winback Zone attribute), you may want to exclude them from certain activities.
There are two ways to use the outputs of Zaius' Customer Winback:
Winback Campaign Recipe
By clicking the "Install Winback Campaign Recipe" button at the top of the page, you will install a pre-built campaign. The winback campaign configuration includes a specially-targeted segment and an email campaign that sends three touchpoints over two weeks. When it is installed, the campaign draft is created in your account.
The segment is designed to capture customers who have just crossed into the winback zone each day. This is the optimal time to re-engage, when your efforts will have the highest reward. Each day, a new set of customers will cross this threshold. Only those new customers will show in the segment, and they will then start in the campaign flow. To use the campaign, follow the steps below:
- Add additional enrollment rules or exclusions, such as excluding those who have gotten a previous winback campaign.
- Select a start time. This campaign should be set as Recurring and not Continuous, as the winback data is updated once per day, overnight. Choose a time of day to send the campaign each day.
- Update each touchpoint to include your brand’s messaging, headers, and footers.
Winback in Segment Builder
You can also use winback insights directly from Segment Builder to create your own campaigns, analyses, and segment sync.
The Winback Zone attribute takes on the values of Engaged, Winback, and Churned, based on the customer's probability to still be a customer. The Winback Type attribute takes on the value of New when the customer is in the winback zone and has just been added to that zone the last time the model was run (overnight). The customer will no longer have that value the next day. This value helps you target ongoing campaigns targeting only customers new to the winback zone.
How the Model Works
The major input into the model are event days for your brand's customers. The events themselves are not considered, but, rather, each day is a "hit" or "miss" depending on whether the customer actively engaged with the brand on that day (opened an email, visited the site, ordered, etc). The more sources of data you have integrated with Zaius, the clearer picture of your brand's patterns of engagement the model will be able to consider.
Once the model is built, customers are scored on their pattern of events to determine the most likely path they will follow between days since last engagement and probability to still be a customer. Based on the modeled path and their current days since last engagement, the customer is assigned a probability of still being your customer, which translates to being in the Engaged, Winback, or Churned zone.
For the data scientists
The model starts with the Pareto NBD (negative binomial distribution) model, which is a well-accepted customer lifetime value model. While the original model predicts purchase events, the Zaius model predicts any kind of engagement. This model uses the non-subscription version of the Pareto NBD, essentially assuming that future interaction is not unduly influenced by a past interaction, such as in a subscription model. Zaius validated that model is predictive of future interaction by making historical predictions and then looking at subsequent re-engagement.
The winback opportunity starts at the measure of liveness where increasing liveness has a larger corresponding increase on future engagement; this starting point is different for each client. The winback opportunity ends and a customer is considered churned when increasing liveness has a smaller corresponding increase on future engagement. Each section (Engaged, Winback, and Churned) has an aggregate likelihood of re-engagement with all activity held constant.