Practically every experienced marketer today knows how to calculate customer lifetime value (CLV).
The specific CLV formula may deviate slightly from one brand to another and across different niches and industries. The general method for calculating customer profitability for specific individuals and segments, though, is essentially the same at just about every organization operating today.
Some marketing teams today calculate CLV with — and, in turn, make critical business decisions from —manually updated spreadsheets fed into a model, using some of their customer data sourced from disparate, siloed systems in their stacks.
These aren’t exactly the ideal means to evaluate existing customers and the sales figures associated with them: customer acquisition costs, total revenue earned over time, average customer order value, and other important metrics that paint a picture of their overall value to one’s organization.
In fact, there’s a far more efficient means to not only calculate customer lifetime value as it stands today; but to enable you to use CLV in conjunction with everything else you know about a customer.
How to calculate customer lifetime value: A refresher
As noted, there are distinct customer lifetime value models used at companies worldwide, but the general customer lifetime value formula (not to mention the formulas for calculating retention and churn rates) is the same at just about every business you’ll come across today:
In short, the bare-bones customer lifetime value definition is the revenue an organization can expect to generate from a given buyer or subscriber in the long run.
Purchase history for specific products and services and online and offline customer behaviors factor into a marketer’s ability to measure and steadily improve CLV over time.
What’s more, CLV informs other departments within a company as well: from who B2B sales representatives decide to nurture to how B2C customer support teams prioritize ticket responses.
All in all, customer lifetime value represents arguably the most important metric for businesses of any kind and in any industry today — and one that requires constant monitoring by marketing.
Why predictive modeling in a CDP matters
Knowing how to calculate customer lifetime value is essential today. But doing so in a quick and efficient manner that allows for measuring customers’ profitability in real time is even more vital.
That’s where predictive modeling and a CDP come into play.
The role of artificial intelligence (AI) in marketing programs today continues to grow at a remarkable pace. But it’s easy to see why the technology — and, specifically, machine learning — is being leveraged more frequently: It helps everyday marketers automatically calculate predictive scores like CLV based on a vast volume of data.
What’s more? In a customer data platform, you can easily calculate and attach CLV scores to an individual level profile to build segments based on high or low CLV customers, for example. With data coming directly from a profile database, you don’t have to worry about pulling manual lists before building CLV models.
With AI Workbench, BlueConic users have access to out-of-the-box, machine learning predictive models that can tell them a particular buyer’s propensity to purchase, when a subscriber may churn, and — that’s right — the customer lifetime value for specific individuals.
By incorporating a customer’s historical transaction data, AI Workbench constructs a predictive model that calculates CLV based on:
- Probability a customer will buy again: With historical purchase information and behavioral data (e.g., frequency, recency, and intensity of digital activity) incorporated in your CLV model, you can accurately gauge which customers or clients are likely to buy or subscribe again down the line.
- Likely number of purchases made in the prediction period based on past buying behavior: Similarly, the machine learning predictive models deployed in AI Workbench can use the same data sets to determine how many times an individual or segment of customers are likely to buy over time.
BlueConic customers can, in turn, get a CLV score that they can attach to individual-level profiles. By enriching profiles with predictive scores, you can get the customer insights you need to not only retain customers but also cross-sell and upsell them with personalized and individualized offers; or to create bespoke experiences based on their lifetime value.
What you can do to boost CLV scores and customer loyalty
Now, how you elevate the average CLV score for all customers — and, in particular, those for your highest-value buyers — will certainly differ from the strategy other companies employ.
If a furniture retailer aims to increase customer lifetime value for those who order $1,000-plus of home furnishings once every 12-24 months, its predictive CLV model for those buyers will be different than one for a retail business whose customers buy lower-cost items (e.g., those less than $50) more often.
Why is this the case? Because the former company isn’t likely to get business from an individual making a sizable furniture purchase more than once a year, in all likelihood, whereas the latter organization may see repeat customers return to buy again within weeks or even days of their most recent purchase.
Factoring in your business model with your CLV formula
As BlueConic’s Customer Success Manager, Ali Paradiso, noted on our AI-for-marketing webinar, you can tweak your CLV model to match your sales cycle and your products’ and services’ price points:
- If your brand has a typically long sales cycle and low-frequency products, your predictive customer lifetime value model in AI Workbench, for instance, will likely require a longer timeframe in order to capture data necessary to make an accurate calculation. For example, you might want to use purchases over weeks or months, instead of days for your model. Then, ensure that the chosen observation period is large enough to observe multiple purchases in order to deliver the most accurate calculation.
- Conversely, if your business has a generally short sales cycle and high frequency, you could use daily or weekly purchases to calculate CLV.
The more you understand your distinct business model, the better your machine learning predictive modeling efforts will provide relevant, detailed, and timely insights about your entire customer base you can use to continuously enhance your overall customer lifetime value.
Creating new customer segments based on CLV models
Once you have a wealth of info about your customers’ value to your organization — order frequency, purchase price points, propensity to buy, etc. — you can develop dynamic customer segments based on CLV scores and other key data points to help with cross-channel personalization.
It’s no secret there are far greater marketing ROI opportunities by targeting or creating premium experiences your most valuable customers — repeat buyers, long-time subscribers, and luxury-product purchasers – than your least valuable customers: one-time buyers who’ve shown little intent to buy again (or at least in the near future).
With CLV scores that you can schedule to update in BlueConic profiles and segments that update in real time, you will always have a clear distinction among your audience groupings, be able to prioritize your efforts, and ensure delivery of the best messaging or offers to each segment.
In recent research, Salesforce discovered retail shoppers who receive personalized recommendations via brands from whom they’ve bought from before accounted for only 7% of their overall call-to-action clicks. However, those recommendations drove one-quarter of orders and revenue for said deals.
If you know the previous buyers who seem most likely to purchase from you again and are able to target said individuals with pertinent offers at opportune times (e.g., following their most recent site visit), the cumulative customer lifetime value for those individuals is likely to bump up appreciably.
Time to prioritize advanced customer lifetime value models
If you have martech in place that helps you track customer lifetime value and other core marketing metrics for which your primarily responsible, that technology can likely give you some useful insights about who the most prized buyers are in your database and deserve your marketing attention.
But predictive modeling through a solution like ours can help you and your team extract even more value from your most important buyers (and perhaps even the lesser-priority ones for whom you can test discounts and deals) and interact with them in real time as their behaviors change.
Simply put, by incorporating machine learning functionality into a CDP like AI Workbench, you can uplevel your targeted marketing activities and digital strategy as a whole.