The Crawl, Walk, Run Guide to Lifetime Value

If understanding a customer’s long-term value were easy, every marketer would do it.

But measuring LTV (long-term or lifetime value, referred to in some circles as customer lifetime value or CLV) is extremely tough. Brands must be able to identify and track customers over time, and they must work for companies that care about retaining long-term customers, not just making good numbers next quarter.

Because of these challenges, few marketers measure this stat comprehensively.

Yet, LTV – which measures beyond clicks or single purchases – is a better gauge of brand health than short-term metrics like click-through rate, which are like junk food compared to the nutritious meal LTV provides.

“I don’t think there is any measure out there which is more predictive of your company’s eventual success than customer long-term value,” said Anant Mathur, global head of analytics at Essence. “It’s a proxy of how well you retained your customers and how satisfied they were.”

Brands make better decisions when they use LTV, which saves them money.

“You will acquire more valuable customers, and will allocate dollars more effectively,” said Matt Greitzer, co-CEO of Amnet. “And you will have a better understanding of the trigger points that drive people along the path of acquisition, retention and ongoing loyalty.”

Brands also avoid spending marketing dollars on one-time customers. “You may be spending money on customers who will never come back to you,” said Michael Hemsey, president of Merkle Loyalty Solutions.

The value a brand places on LTV depends on the product it sells. People use razors for decades, while diapers are more of a life stage product, said Christine Peterson, managing partner and digital investment lead at Mindshare.

Figuring out lifetime value has often fallen to consultancies, who would measure profitability by segment, said Ram Singh, head of analytics at Performics US and a former management consultant. These analysts might look at long-term value as a measure of risk, not just as a way to optimize media. A product where it’s easy to switch to a competitor, for example, might need to lower its LTV calculation.

Using LTV calculations to inform media decisions is spreading, but still not widely used.

“While people understand it, LTV is not common currency in how performance-based campaigns are managed today, which is detrimental to the industry,” Greitzer said. “There is a real opportunity for people to up their game in this area.”

Crawling: Starting With The Quick Win

Brands who want to measure LTV can start by looking at metrics that approximate long-term value.

For example, they can look at repeat purchasers over a short period of time, like a 90-day cookie window, said Jason Colon, regional director of commerce for Resolution Media.

“Instead of looking at a single purchase attributed to a media touchpoint, you can start to expand that out and look at multiple purchases driven after that touchpoint,” Colon said.

Once enough data has been collected, start bucketing users based on LTV.

“Crawl is differentiating between high-, medium- and low-value customers, and making investment decisions with that in mind,” Greitzer said.

Brands should look for metrics that reflect who their best customers look like, said Mathur. Does the brand care about the ability to upsell or cross-sell current customers? Frequency of purchases? Total dollar value of orders?

“The organization has to figure out how it measures success,” Mathur said.

Brands interested in harnessing LTV need to start building out their data stack, including tools for data storage and analysis – all of which need to be hooked up to marketing execution solutions.

“You need to be able to not just store the data, but also act upon it,” Mathur said. That means ERP and CRM systems as well as a DMP and DSP for media execution.

And brands need to staff up, because good analysis requires both data scientists and people who can make that data actionable. Brands should scope their agencies accordingly.

“Hire talent that understands the data, but can tell stories in a way that make it compelling to decision-makers,” Mathur advised.

Finally, brands need top-level executives to be on board – a challenge during a time CMO tenure averages just 42 months.

“It needs to start at the C-level,” Hemsey said. “What we are talking about is an enterprise strategy. Because [LTV] touches every aspect of the business, from marketing, digital and finance, they all need to be working against a common goal driven by the C-suite.”

Peterson agreed. “It needs to be embraced from top to bottom at clients, and that’s a difficult shift because of the way people are compensated,” she said.

Walking: Optimizing Everywhere

Once brands prove the results of using LTV, they can make the case that the metric should factor into decision-making.

“[Brands] are starting to embrace a blend: delivering on quarterly sales, but not at the expense of a deeper investment in creating that loyalty and lifetime value,” Peterson said. “Immediate sales impact isn’t the only piece of the puzzle.”

With executive approval, brands can start optimizing campaigns to get more valuable customers, not just the ones that will bring in the most value on their next purchase.

Feeding LTV into programmatic campaigns, for example, allows for better frequency management.

“Getting into the programmatic space, there should be a different frequency cap set for audiences with different LTVs,” Performics’ Singh said.

Bringing in LTV may also make some media buys look much better – or much worse. Affiliate channels, for example, often attract more transient customers.

In order to maximize this potential, brands need to make sure their acquisition teams (which often lead on programmatic) and retention teams are communicating. This process requires incenting CRM and media teams to share LTV metrics to apply toward common goals.

If the units are separate, the acquisition team won’t be incentivized to find high-value customers. “That is an organizational mistake, to have a separate team with different goals that may not even communicate regularly,” Greitzer said.

“Marketers need to push harder on this collision of CRM and media that’s happening to understand your customers better and make your media more targeted,” Colon said.

Because the best place to grow LTV is by reaching existing customers, the CRM team needs to be intimately involved. Email marketing, for example, is a great place to increase the LTV of existing customers.

“Your best customers are ones you already have,” said Amnet’s Greitzer. If brands encourage a customer make a repeat purchase, or increase their order value from $50 to $75, they improve the LTV metric.

Running: Making LTV Predictive 

For brands to run, they need to make their LTV metric more ironclad.

“To run, you need to go from a historical perspective to a predictive one,” Mathur said. “The truth is that the people who were most valuable in the past may not always be predictive of what’s going to happen in the future.”

Good measurements of LTV take into account what someone is buying, with higher weight to customers that buy low-value items frequently and a lower weight to one-time big-ticket customers.

“I think of long-term value as a dynamic statistic that will change over time – and it might be more vulnerable to changing depending on what basket [of goods] is being considered,” Singh said.

Forecasting, however, requires advanced tools, not all of which are ready to market. GroupM for instance is working on a predictive model for its clients.

“GroupM is coming up with new ways to attribute value and make predictive value measurements for what kind of impact an investment now would have in 20 years,” Mindshare’s Peterson said.

In reality, this stage is more backflipping than running – it’s so difficult that only the biggest brands typically get to this point.

One reason it’s so difficult is that being predictive about LTV requires looking at how customers will feel about a brand years out.

“Run, in my experience, is rare,” Greitzer said. Only a select group of marketers can be “aggressive in their investments without being beholden to a metric that’s easier.”

Advanced LTV might seem unattainable, but the rise of AI and machine learning will make it more accessible.

“The predictive power of optimizing toward LTV is one that is nascent today, but we will see more action over the next couple of years,” predicted Amnet’s Greitzer.

Essence’s Mathur agreed. “This is a classic machine-learning application,” he said. “And this will happen faster than most people think.”

Another trend that might make LTV calculations more commonplace: the rise of consultancies focused on executing clients’ strategies and vying for contracts with brands.

“Consulting companies are making acquisitions in data or media management,” Singh said. “They are not advising the clients and moving out, but going into implementation.”

The cozying up of mar tech and ad tech will drive further use of long-term value, Mathur predicted. “When you speak to an individual holistically, you drive higher customer long-term value.”

1 Comment

  1. This is a great breakdown of how a business can start incorporating LTV into their business using "baby steps." At Zodiac, we make it easier for businesses to get to the "running" stage because our statistical models require only historical transaction information to produce long-run predictive insights on a firm's existing customer base. It is absolutely correct that historical behavior does not usually continue into the future (or as we say, customers tend to get worse over time), which makes it important to use proven statistical methods to model that behavior correctly so you don't end up over-spending to acquire customers that will make one or two purchases and then churn.

    Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>