GROW: Margin-Based Marketing 2.0
Leveraging the Power of CLV to Get the Most From Your Marketing Investment
By Giorgo Paizanis, Lead Growth Architect at BCG Digital Ventures; and Liz Melton, Client and Partner Ops Manager, and Caroline Kratofil, Marketing Manager, at Retina
After our first piece on Margin-Based Marketing, LA-based startup Retina got in touch with some valuable perspectives. This co-authored piece expands on the power of customer lifetime value (CLV) and how it can be applied to MBM.
The global COVID-19 pandemic and subsequent lockdowns have compelled companies to rethink how they find and serve customers. According to a recent study from BCG, 75% of executives say the pandemic has upped the need for digital transformation. To adapt and thrive in the future, most companies will need to evaluate their marketing stack and make sure the right tools and methods are in place to win in this new reality.
What’s wrong with the everyday marketing strategy?
Today, many marketing teams segment their customers based on the products they buy. In this scenario, marketing teams will pay more to acquire customers who either purchased expensive products in the past or have leading indicators that they may buy pricier items. There is a clear downside to this cut and dry approach.
Some customers who buy expensive items eventually disappear, providing little future revenue. Other customers initially buy mid-tier items but stick around longer, generating more revenue over time. What is more, some customers buy the cheapest product first, get hooked, and end up being among the most profitable customers in the long run.
The problem is, marketers don’t know how valuable each of their customers is一they just use a proxy. But imagine if you could project every customer’s lifetime value up to 15 years into the future. How might that kind of information influence your marketing strategy?
Introducing Margin-Based Marketing (MBM) and Customer Lifetime Value (CLV)
While the shift to digital brings many benefits, it also intensifies competition and flattens the market, as everything is just a few clicks away. To successfully market a product or service on digital platforms, it is important to focus on what we call Margin-Based-Marketing (MBM): “the practice of optimally allocating marketing capital to maximize profit or achieve a target profit margin.”
With the right tools in place, you can track Customer Acquisition Cost (CAC) and Customer Lifetime Value (CLV), enabling you to focus on generating the greatest margin per dollar invested in your marketing program. Not only does this data give you valuable feedback about your performance, it can also help reduce the cost to acquire new customers, increase the proportion of high value customers, and extend lifetime values.
Linking marketing inputs and bottom-line results
The first step to implementing MBM is to link marketing inputs to bottom-line results. There are two primary ways to do this:
- Attribution modeling
In digital marketing, attribution models are used to identify a set of user actions that contribute in some manner to a desired outcome (e.g. sign-up or purchase), and then assign a value to each of these events. - Marketing Mix Modeling (MMM)
Precise source attribution is challenging to define in offline media such as TV or out-of-home media. In such cases, the impact of a particular marketing input is inferred through econometric analysis techniques and (ideally) hold-out experiments.
You can read more about marketing measurement here.
With sources attributed, we can get a much clearer line of sight from our marketing allocations to the value generated by that investment.
Data collection, processing, and analysis
There are several methods to calculate customer lifetime value. Before we get into how to calculate CLV, let’s first make a distinction between historical CLV and predictive CLV.
Historical CLV is the actual value generated by a customer since they first purchased from your business. While this is an important measure, it does not give the complete picture for a few reasons. First, the historical CLV is only possible to calculate if you have a long track record. Even if you do have some historic data, it may not be enough to show the full lifetime (e.g. those customers may purchase again). Lastly, the historical approach will tell you how the customers you acquired months or years ago performed, which is not necessarily how the ones you acquire today will perform. Therefore, when talking about CLV we refer to predictive CLV, which is a prediction of the future lifetime value of a customer acquired today through a particular channel.
The value we calculate and assign to customers may also be net income. In other words, it is important to go beyond revenue and remove variable costs such as cost of goods sold (COGS) or the cost to serve (e.g. sales/support). If the average customer lifetime is particularly long (e.g. several years), it is wise to discount future years to account for the time value of money.
The lifetime portion of the CLV equation can be challenging to compute. To analyze lifetime we must track users on an acquisition cohort basis, meaning that we will group data not by the transaction date, but by the date that the customer first purchased. With this approach we can observe the retention curve and from there derive the average expected lifetime for a newly acquired customer (you’ll find some examples of cohort-based data analysis below.)
In addition to tracking CAC, CLV, and Gross Customer Margin, each of these metrics can be further broken down into fundamental inputs. For CAC, these inputs could be cost per impression, click-through rates, and purchase conversion rates. For CLV, the inputs could be: average lifetime duration, retention curves, average order size, transactions per period, margin per transaction, etc.
Leverage predictive analytics and automation to do this at scale
While it is possible to calculate the metrics outlined above in a simplistic and aggregated manner, either for an entire customer base or for high-level groupings (e.g. acquired through paid channels vs. arrive as organic and direct traffic), these simple methods are less actionable. The ideal setup will be bottom-up in nature, allowing for individual customer-level predictions that can be segmented in myriad ways to provide actionable insights.
For example, it is customary to look at CAC and CLV on a segmented basis by acquisition channel to influence marketing spend allocation decisions. Another option is to segment customers based on their purchase patterns and behaviors, which is more useful in retention and lifecycle marketing activities.
Implementing these strategies at scale requires both user-level data and advanced predictive analytics and automation techniques. Thankfully, over the past decade we have seen a plethora of third party tools enter the market to make it easy for even the smallest firms to implement MBM.
Taking advantage of predictive analytics
There is a clear need to increase efficiency and effectiveness of digital marketing. Companies must drive greater CLV through improved targeting and lifecycle marketing to keep up with the competition. Organizations are increasingly turning to sophisticated modeling techniques that help them calculate individual customer lifetime value.
While “Buy ’Til You Die” models are very popular, the best CLV models specialize in predicting the most accurate customer lifetime value as early as possible. Retina AI, for example, specializes in calculating Customer Lifetime Value at or before first purchase. Accordingly, their clients have revolutionized their approach to both acquisition and retention marketing. Armed with CLV, they:
- Create lookalike audiences to specifically attract high-value individuals
- Allocate their CAC based on the long-term value of each prospect or customer using value-based bidding methods
- Discern what kind of content is resonating with their highest value customers and refine their messaging
- Revamp their retargeting campaigns to convert only the highest value prospects
- Determine which customers should or should not receive discounts or promotions based on their CLV and purchase behavior
What should you look for?
Look for companies that have integrations with Customer Data Platforms (CDPs) like mParticle and Segment. CDPs collect, standardize, and control the customer data that powers Retina’s early CLV and Personas models. These integrations will enable you to optimize media spend on lookalike audiences, pick the most relevant keywords, create customized user experiences, and identify the offerings and product bundles that resonate most with high CLV customers.
Additionally, consider vendors that can predict at first transaction and over the long term. Therefore, the CLV you use for your marketing strategy will be resilient to macro and micro trends, unforeseen events, and seasonality (like COVID-19, holidays, etc.). Next, the model should be interpretable. You should understand exactly what behaviors are driving customer lifetime value and how you can incorporate those into your campaigns.
You also need CLV at the individual level, not in aggregate. Aggregate CLV simply does not allow you to adjust who you’re targeting with re-engagement or acquisition efforts to maximize the benefits of each campaign. A good CLV model assesses the commonalities of all the customers unique to your business, then combines that information with per-customer behavior in order to predict future purchases and future spend with your brand.
Lastly, you want a model that has less than 20% individual error and less than 5% aggregate error. With these requirements, you’ll be confident in predicted CLV across any time span.
How do you take action on CLV?
Acquisition
Once you’ve calculated CLV, you can use it to determine your target customer acquisition cost. Instead of simply targeting customers that bring in high initial revenue, focus your targeting and acquisition efforts on high CLV customers. Even if some of their first purchases might be small, they will bring your business more value in the long run.
More specifically, a customer acquisition is unprofitable only if their acquisition cost exceeds their lifetime margin — not just the margin of their first purchase. For most businesses, it makes sense to keep a CLV to CAC ratio around 3:1 for each marketing segment. Spend too much (like a 1:1 ratio) and acquiring these customers won’t be profitable. However, spend too little (like a 7:1 ratio), and you’ll be missing out on profitable customers whose acquisition cost is above your current bid cap.
Retention
In the course of their customer journey, many customers will feel unsure if they will purchase from your business again. In that key moment of indecision, successful customer outreach can pivot a customer’s journey from the possibility of churning to long-term retention. However, placing that exact moment is often left to guesswork.
You can use customer-level churn predictions to identify customers who are just starting to be at risk of churning. Reaching out to those customers at exactly the right time boosts retention far more effectively than sending an impersonal mass email.
After calculating the lifetime value of each of your customers, you can unlock the following three strategies for better marketing:
- Value-Based Lookalike Audiences
- High/Low Value Audiences
- Value-Based Bidding
Connecting your CLV data to ad platforms can be done manually or via a Conversions API. This API is a practical way for advertisers to pull in external data to create custom audiences in the platforms. Under the Data Sources tab in Facebook Events Manager, you can choose your partner under Settings. Facebook currently partners with a variety of customer data platforms, system integrators, attribution platforms, and more.
Who has done this before?
Madison Reed, a leading haircare and color brand, understood that specific customer attributes drove higher lifetime value, but struggled to identify those characteristics. They needed to find the right data points and products that positively impacted CLV and product-market fit.
By the same token, they also knew that hiring more data scientists was not feasible from a cost or timing perspective. At least two data scientists would’ve been required to build state-of-the-art models from scratch, run and maintain pipelines, and determine new ways to apply CLV in Madison Reed’s business. Plus, it would take a significant amount of time for HR to headhunt and then onboard new data scientists. By hiring Retina, Madison Reed was able to gain this expertise at a tenth of the cost.
Madison Reed leveraged Retina to group customers into personas based on customer lifetime value. With CLV-based segmentation, Madison Reed discovered that some of their products were not resonating well with their high-value customers. By deprecating those products in favor of those with higher product-market fit, the company saw a 23% increase in average CLV.
Moreover, Madison Reed started identifying unprofitable customers and restructured their marketing strategy to employ value-based lookalike audiences on Facebook. As a result, Madison Reed experienced a 2.2x lift in high LTV conversions, a 34% lift in ROAS, and their LTV to CAC ratio grew by 150%.
Conclusion
Unprecedented circumstances in 2020 led many companies to accelerate digital transformation strategies and focus on margin-based marketing. By shifting goals to increasing customer lifetime value and decreasing customer acquisition costs, companies are better positioned for success in 2021 and beyond.
To get started with customer lifetime value, it’s critical to choose a model that is accurate at the individual and aggregate level and one that provides predictive CLV metrics early in the customer journey. With this foundation, marketers can implement CLV at scale with strategies like lookalike audiences and value-based bidding to improve acquisition, retention, retargeting, and more.
About the authors:
Giorgo Paizanis is a Growth Architect at BCG Digital Ventures, a corporate investment and incubation firm. We invent, build and invest in startups with the world’s most influential companies.
Liz Melton and Caroline Kratofil are Client and Partner Ops Manager and Marketing Manager at Retina, a customer intelligence partner that empowers businesses to maximize customer-level profitability by focusing on early customer lifetime value.
