CLV Marketing: Using Customer Behavior Analytics & Retention Rate To Improve Profitability

Your customer data should be telling you more than numbers. Customer Lifetime Value models take data such as repeat purchase rate or average time to repurchase and turn them into incredible insights for how to find the right customers, increase purchase frequency, and maximize your eCommerce success.

How do you visualize your business model? What about your customer journey? And how do you unite the two and attempt to model how they interact to know how you can grow your business efficiently? The answer is suprisingly not that complicated.

Marketing Funnel Modeling for Acquisition, Conversion, and Retention

In recent years, Customer Lifetime Value has been rising to the forefront of the finance and marketing worlds. This newfound preeminence has in large part been brought on by the likes of Peter Fader and Daniel McCarthy, whose work in CLV has led to the emerging framework of Customer-Based Corporate Valuation. This analysis allows for a more granular assessment of a company’s customer base and a bottom-up approach to forecasting cash flows relating to overall business valuation. For a more detailed explanation on this completely new way of valuing businesses, the Harvard Business Review wrote an in-depth piece in early 2020. CLV isn’t just a valuation technique though–it’s a way of bringing business data to life.

What does lifetime value mean for eCommerce companies? By turning the overwhelming volume of customer-level data into a rich story of the customer journey, e-retailers can fine-tune their marketing efforts to cater to their most valuable cohorts. Plus, a solid understanding of CLV is imperative for making decisions related to customer acquisition costs. Most eCommerce companies know what their average order value is and how much they are spending on paid platforms, but measuring acquisition costs off just one purchase will never reveal the whole story and is a surefire way to miss out on key behavior insights.

Tadpull’s cross-channel marketing strategy is designed around our marketing funnel framework.

We like to use all of the resources we can in order to connect with a customer, cultivate a relationship, and eventually close with a transaction. Identifying all of the interactions a user has with your company before a purchase is crucial.

The customer conversion funnel


However, while acquiring new customers is essential to the growth of a company, it’s only half the battle. Nurturing relationships and turning one-time purchasers into repeat customers will create long-lasting cash flows and high value.

The efficacy of that retention can be measured as Customer Lifetime Value, which predicts how much your customers will be worth according to their future purchases. Showing growth in this lifetime value can give reassurance that a company has a strong base of repeat customers.

Understanding the behavior of your repeat customers is incredibly powerful. At Tadpull, there are four main ways that we try to capitalize on this data:

  1. CLV model basics: create a retention timeline and map your best customers
  2. Drive growth by continuing to increase the value of those customers
  3. Understand and optimize CAC
  4. Find customers that have similar characteristics to your highest CLV customers

What A CLV Model Should Include

Average Time To Repurchase

Understanding when your customers are likely to purchase can lead to more effective and targeted marketing efforts. One important factor to consider when developing a strategy to retain customers efficiently is customers’ average time to repurchase. Different customers will have different times between their purchases, but there will always be trends in purchase cadence that will inform what kind of customers they are. 

For example, most of our clients see customers’ time between purchases decrease as the number of repeat purchases increases - the boxplot below shows these trends.

Time between purchases boxplot


While there is a huge range in repurchase cadence, there is a consistent decreasing trend among the median time between purchases. This can be one of the variables that help us identify customers that are at risk to churn.


Recency, Frequency, & Monetary Input Models

While investors have to rely on publicly disclosed data to calculate a firm’s unit economics, smart marketers can take advantage of their own data. The sales lines sitting in your ERP system can be a gold mine of actionable data when used correctly. These datasets can be used to find out how often a customer purchases, how long they have been a customer, and how much they spend. Conveniently, those metrics make up the three inputs needed for a recency, frequency, and monetary (RFM) CLV Model. 

To dive in further, building a Customer Lifetime Value model based on customers’ recency, frequency, and monetary value of purchases gives us the opportunity to assign a probability for when they may purchase again. If we see a customer who is past their average time to re-purchase and their probability has decreased, it could be a great opportunity to send them a personalized email with a coupon code or other deal. 

The goal is to not waste time and money engaging with a customer that is already planning on purchasing, but to focus on those who may be on the fence.


Knowing If Customers Are ‘Dead’ or ‘Alive’

These models are also highly dependent on the probability of a customer being ‘alive’ or ‘dead’. If a user has historically purchased monthly, but has broken that pattern over recent months, our model can assign a probability to whether or not they are ‘alive’. Figure 1 shows this concept in practice. We see that this customer’s probability of being alive immediately decreases following a purchase, but returns to the previous starting point at the time of the next purchase. Over the long run, a customer that purchases more will have a higher initial probability of being alive after they repurchase.

Chart of active customer likelihood


Given a customer’s probability of being alive and the value of their purchases in the past, our model is able to return an expected value of this customer’s future purchases. 

Maximizing Value From High-CLV Cohorts

With our models to drive growth in place, we can start to plan new strategic paths to profit by increasing the value of high CLV customers. How do we effectively build that relationship and create loyalty? We don’t want to annoy a customer by reaching out too often, but we also don’t want to miss out on their full potential. No user wants to be hit with a dozen emails, display, and Facebook ads concurrently.

You could serve them with special offers, early access, or other incentives to be rewarded for their loyalty. You can also be sure to track product reviews from them, or if you’re gathering NPS comments or any other form of customer feedback, make sure their comments are taken into consideration (or even replied to with email automations). But deciding the best strategy to use is oftentimes difficult and overwhelming.

The method to this madness can be found in customer lifecycle marketing. Customer retention is all about understanding your customers and developing a strategy to engage them at the right times. 

Strategies To Retain New Customers

If a customer has purchased (and had a good experience) we should hope to keep in contact with them. Hopefully, they subscribed to our email newsletter and our brand can stay top of mind with our periodic emails. Along with normally scheduled sends, it is always a good idea to send segmented emails to these customers as well:

  • Are they yet to purchase this holiday season when they had already purchased at this time last year?
  • Did they buy a product that should need a replacement or complementary product?
  • Do you have SKUs with broken sizing that could incentivize an early purchase?

The VIP treatment is also a great tactic for many customers. Email is a great way to send special deals or early access, but SMS is another option to engage with your most valuable customers! In fact, one of our clients saw a 40% click rate and a 6.7% conversion rate for a Back In Stock SMS campaign (compared to 3% and 2%, respectively for a comparable email campaign).

Strategies to Reactivate ‘Dead’ Customers

As we see customers slip past their expected repurchase times (and identifying this is huge in itself), there are several ways we can try to prolong the life of the customer. From email to Facebook or Google Display campaigns, it is generally worth fighting to get that customer back.

Win-back campaigns are most effective when they are personalized and often involve aggressive offers to re-engage; but that’s OK! Comparatively, the cost of offering a discount is usually less than the acquisition cost of a new purchaser. If they don’t take advantage of the offer, nothing is lost–but if they do, we are winning back a loyal customer.

All in all, there are a lot of pieces that go into a strong customer base. We always strive to acquire new customers for our clients - and believe this is a good measure of a company’s health - but the majority of revenue will often come from a subset of loyal customers. This isn’t a bad thing, especially if you can nail customer lifecycle marketing. 


Optimizing CAC For Better ROI

What if you could identify which customers are more likely to purchase multiple times and spend the most money over their lifetime?

During the process of developing a strong understanding for how much customers are worth over time, it’s necessary to consider our customer acquisition costs. Specifically, the best area to focus on is the CAC (Customer Acquisition Cost) to LTV ratio. The CAC-to-LTV ratio can give you a better understanding of your customer base, as well as how acquiring new customers will affect your future cash flows. 

An example: A client we work with operates in the highly competitive high-end apparel space. Their products are expensive and their name is unknown, but customers need to be acquired. If they sell a $100 t-shirt at a 30% discount, plus it costs $50 to produce the shirt, how much room does that leave in the budget for them to spend on acquiring that customer? If you’re only looking at the one purchase above, not much. But if you look at the fact that the average lifetime value of a repeat customer is over $500, now they can free up more room for acquisition costs if needed.

If companies can value the total profit that a new customer will statistically contribute in their lifetime, then all it takes is some quick math and presto–Paid channel managers are able to optimize their spending in a much more scalable fashion. 


Find More of the Right Customers Using CLV Models

The biggest impact that a robust CLV model has is that it allows businesses to find users who display similar characteristics to their current highest lifetime value customers. At Tadpull, we have seen massive success for clients when marketing to lookalike audiences based on CLV data on Facebook or when informing audience targeting decisions in Google Ads based on high CLV audience attributes. 

High CLV customers can also send signals about what acquisition channels are the most valuable. If you find that most of these loyal customers came through either Facebook or email, it may be a good idea to expand your Facebook reach while more aggressively grow your email list. It may also be advisable to work on identifying any confounding variables that led them to those channels for a more complete understanding of their customer journey.

Customer Lifetime Value can have big implications for any business, like valuation, optimizing marketing costs, or simply having a better understanding of your customer base (after all, knowing how many loyal customers you have is an important metric in itself).

But CLV isn’t just a powerful concept because it will help you make better marketing decisions. It’s also important to note that it can only come from having one unified way to examine the multiple data sources that every eCommerce business has. Even though sales order data might be sufficient to develop the start of a predictive CLV model, the real magic happens by leveraging all of the data at your disposal–from users’ site behavior to creative/channel performance–in order to create a holistic view of our customers’ journey and path to purchase with your brand. Tadpull’s software Pond was created for exactly that purpose, and you can learn more about how it revolutionizes eCommerce analytics here.

Overall, there are innumerable different strategies for connecting with your customers at the right time. Many eCommerce sellers rely on ‘gut feelings’ to determine the best ways to retain their customers without taking advantage of their data and consequently miss out on huge growth opportunities. 

If you’re interested in a consultation with our growth and data science experts to learn how you can best programmatically reach your customers, reach out and drop us a line!

Image: freepik