Mitigate churn with predictive insights and custom CLV

Spend your finite resources on retaining the right customers

Retain customers. Reduce churn. Grow the business. Simple enough, right?

As a marketer, you know that is far from the truth. In an increasingly complex market where customers have more options and have mounting expectations around customer experience, it can be really difficult to retain customers. As we know from general marketing principles, the wider the net you cast, the less effective your efforts are in the end. Simply put: you cannot retain every customer.

So who can you retain? To answer this question, you canuse the customer data that you already have to make predictions about which customers are likely to have the highest customer lifetime value (CLV).

Klaviyo’s built-in predictive analytics combined with the custom CLV prediction window allow you to predict future behavior with confidence. You can answer key questions like:

• Which customers are likely to place another order?
• Which customers are at the highest risk of churning?
• When can I expect a given customer to place their next order?

With reliable insight into future behaviors, you can intervene proactively to get more conversions and keep your brand top-of-mind for customers who are likely to have the greatest long-term value for your business.

Note: The custom CLV setting is only available for Marketing Analytics and Advanced KDP customers. If you are not on either of these plans, head to our billing documentation to learn how to purchase them.

Align your prediction window to your business cycle

Your customer’s buying patterns are heavily dependent on the type of product you sell, and the nature of the industry you are in. Our standard CLV metrics are calculated based on a set window of 365 days. If you operate a highly seasonal business where customers tend to purchase just once a year, then having a prediction window of 365 days would make a lot of sense.

But what if you want to contextualize the predictions within a different time frame? The custom CLV setting in Klaviyo's Marketing Analytics and Advanced KDP allows you to adjust the prediction window for all predictive CLV values in Klaviyo, so your targeting efforts can more accurately reflect the reality of your business. When choosing your prediction window, consider the following:

‍

• How long your average buying cycle is
• The average lifespan of a customer
• How long your product is intended to last

‍

Develop marketing tactics around predicted metrics

Each prediction metric in Klaviyo offers unique benefits when it comes to proactively surfacing content to customers. These metrics are calculated using all historical data that exists in your account, combined with the custom CLV prediction window that you specified. Explore each of the predicted metrics below to learn what kinds of experiences are best powered by each data point.

Predicted CLV Total CLV Predicted number of orders Predicted gender Expected date of next order

Predicted CLV

What: the amount of money a customer is expected to spend within the specified prediction window

Use cases:

• Send targeted invitations to join rewards program based on expected future spend
• Send “refer a friend” requests to customers with a high value
• Up-sell/cross-sell more expensive products to customers with a higher value
• Entice customers with a lower value to try a product free

Total CLV

What: the sum of a customer’s historical CLV and their predicted CLV

Use cases:

• Use as an analysis metric when comparing the relative value of RFM segments
• Reward customers over a certain threshold with a special perk that encourages brand advocacy

Predicted number of orders

What: the predicted number of unique orders a customer is expected to place within the specified prediction window

Use cases:

• Target customers with a value of 2 or more with the option to subscribe and save on certain products
• Surprise-and-delight customers who are expected to place 1 or more orders with flash sales, special discounts, or offer bundles to encourage more purchases
• Create targeted onsite offers for customers expected to place more orders using the Group Membership API

Predicted gender

What: the predicted gender of a customer; value is assigned likely male, likely female, or uncertain

Use cases:

• Build segments to split your customer base up by gender, and then use the Group Membership API to display hero images on your website tailored to the specific gender
• Feature products in your campaigns that most closely align with the appropriate gender category

Note: This metric is mostly useful to brands that offer gendered distinctions within their products. However, since predicted gender is an estimate, it’s recommended to include some additional product details or images for products specific to all genders.

Expected date of next order

What: a specific date of when someone is expected to place their next order

Use cases:

• In existing flows, use a conditional split where Expected date of next order is within 30 days. Include a personalized product feed with recommendations based on their previous browsing or shopping behavior.
• Create a new date-property flow, triggered a few days before someone’s Expected date of next order. Be sure to include a personalized product feed with recommendations. You could also include a time-sensitive discount with an expiration.

On an individual profile, you can also see a metric called Churn risk prediction. If you want to find the average predicted churn risk of customers in a particular segment, you can include this metric in your segment .csv export. Churn risk percentage can be a useful metric to observe for audience research, but the Expected Date of Next Order metric was created to be a more useful, granular tool that you can use for precisely targeting customers with campaigns and triggering flows.

Key retention plays for keeping your highest-value customers

A successful retention strategy requires a lot more than intelligently-automated communications. In order to keep customers engaged, you need to make sure every piece of the customer journey reinforces the trust that you want to gain from customers. Here are a few key areas to invest in when building out your customer retention strategy: