This post is the second of a series of posts designed to walk startup founders through how we think about Customer Lifetime Value (CLV), how we calculate it (or avoid calculating it), and then how to use it for the benefit of their businesses. If you missed Part 1: The Unexamined Life is Not Worth Living, be sure to check it out.
A common misconception related to CLV is that, if you calculate the lifetime value of each of your customers and plot them on a histogram or probability density functiol.n, they would be normally distributed around some mean value. Something like this:
However, in practice, it doesn’t work out that way. A much more typical distribution of CLV throughout a customer base looksmore like the distribution below. It is generated from data derived from a business services company we work with. We ran a predictive model (more on this in a future post) to calculate CLV and residual customer value, or RCV (see Part 1) for every one ofthe company’s 7,763 customers. Then we assigned each customer to a CLV bin and plotted the number of customers in each bin as a histogram.
Here we see lots of low-value customers clustered on the left of the histogram, and a long tail of higher value customers out to the right. Instead of a normal distribution, it is a power-law distribution. Most of the value of the business derives from the long tail. As the table below illustrates, the 10% of all customers that make up the long tail of customers in the above histogram account for 53% of the expected future value of the business.
More generally, think about this as the Pareto principle or “80:20 rule.” Roughly 80% of value comes from 20% of customers. The power-law distribution of customer value, with the concentration of value in the long tail, leads to an important framework for thinking about your business strategy: Customer Centricity.
Instead of worrying about all of your customers, what if you focus only on the most valuable ones? Customers are not created equal; therefore, they do not deserve an equal share of your startup’s scarce time and resources.The idea behind Customer Centricity is to start with your customers—particularly the most valuable ones—and then work backwards to your product, marketing, sales, customer service, and operations. To do this, first identify exactly who those long tail customers are. Then, study their usage and talk with them to develop an understanding for how and why they behave the way they do. Finally, and align your company around developing products and processes that meet their needs.For a more robust discussion of Customer Centricity, we recommend The Customer Centricity Playbook by Peter Fader and Sarah Toms, whose work is distilled in this section.
What Customer Centricity is Not
Startup founders are naturally inclined to serve the needs of any customer that comes their way. After all, it is hard enough to get people to notice and engage with your business. Thus it’s very tempting to bend over backwards to meet the needs of everyone who finds you and expresses interest. We submit that you probably can’t afford to take this approach.Employing a Customer Centricity strategy helps avoid the pitfall of trying to be all things to all people while operating under the constraint of having limited startup resources. It is not:
- Perfecting customer service
- Aligning strategy with the needs of the overall customer base
- Asking what the “average customer” is worth
- Chasing down product sales to anyone and everyone, at any cost
Instead, Customer Centricity is a mental model and analytical framework that helps you decide which customers to focus on versus deemphasize.
Three P’s of Customer Goodness
As you seek to understand your high-value customers better, Fader and Toms recommend structuring your analysis using the Three P’s of Customer Goodness framework. Three main factors define customer “goodness:”
- Your offering aligns with the customer’s needs
- The customer chooses your offering over a competitor’s
- Propensity: likelihood of being loyal, referring others, buying higher-valued offerings, etc.
- Potential: future value of each customer
Once you start breaking down customer value into these three forms of goodness (among others), you realize just how heterogeneous your customer base really is—and you can better see the opportunity to grow your business by focusing on the right customers.
Customer Centricity in Startups
Customer Centricity applies to large companies and startups alike. We think it is especially powerful for startups because of two factors:
- Achieving Product-Market Fit, or PMF, requires an intense focus on understanding customer behavior. CLV and Customer Centricity are centered on rigorous frameworks to help you build that understanding in an analytical way
- While Customer Centricity helps any company allocate resources throughout the organization, in startups, these resources are often more scarce than they are in more established companies
Early Stage (Pre-PMF): Use RFM and Payback Periods for Marketing Spend
Early stage startups—especially those still assessing the strength of their PMF signals and not yet investing in growth marketing—need to strike a balance between analytical rigor and resource allocation. Such teams almost never have the luxury of an analytics team, or even an individual dedicated to data analysis.If this describes your team, we recommend the following:
- Tap somebody on the team to identify the most valuable customers -- This could be the product leader, the marketing leader, or just whoever is good at SQL and Excel.
- Use RFM analysis (Recency-Frequency-Monetary-value) -- RFM is a relatively easy, lightweight way to identify the customers with the most future value. It’s not necessary or particularly accurate at this stage to quantify long-term CLV, so RFM’s approach of simply identifying the customers with the most future potential is the best we should shoot for. This “good enough” analysis gets you 80% of what you need without requiring a lot of time or training. For more on RFM, .
- Minimize the payback period of marketing spend -- When your company has limited cash runway and does not know enough about future customer behavior, it is not time yet to use CLV to guide marketing spending. Instead, try to minimize the payback period. The chart below depicts how to determine payback period using historical customer value.
Growth Stage (Post-PMF): Build Predictive Models and Use CLV-CAC Ratio for Marketing Spend
Going beyond RFM and actually calculating CLV/RCV in order to drive a Customer Centricity strategy involves more complex data analysis, including the development of a forward-looking predictive model that goes beyond what Excel can do. But a startup should only go down this road if it fits the following criteria:
- A critical mass of repeat customers (at least several hundred)
- A substantial percentage of customers with repeat transactions
- Enough time in market (at least 12-18 months) to build up a transaction history
- An employee dedicated to data analysis, or at least a part-timer/freelancer—more data analytical expertise required
For the companies in our Growth Program with this profile, we encourage them to make use of more advanced predictive analytics. Such models are extremely useful in customer segmentation and developing strategies around marketing spend. Once you are able to calculate reliable customer-specific forecasts for CLV/RCV, you can use that information to target CAC to 30-40% of CLV, as the experts prescribe.
Moreover, because you have CLV/RCV calculations for individual customers, you can segment these customers according to behavioral or profile characteristics and fine-tune CAC for each of these segments. High-value segments justify more aggressive marketing spend.We will go through ways to construct and use a predictive probability model for growth-stage companies and use that information to create behavioral segments in future installments in this series of blog posts.
Since customer value is distributed according to a power law, there is a relatively small group of customers that create most of the value for your company. This phenomenon is the basis for pursuing a strategy of Customer Centricity, which aligns your entire business towards meeting the needs of these most valuable customers. The type of data analysis that you choose to pursue a Customer Centricity strategy depends on the stage of your startup. At the early stage, use the more lightweight RFM segmentation to identify the most valuable customers. With more customers and historical data at the growth stage, you can build predictive models to calculate the lifetime value of individual customers, which will provide a richer view of your customer base.Our next installment focuses on the analysis most appropriate for early-stage startups: RFM segmentation. We call it Part 3: Easily Identify Your Most Valuable Customers.