menu close
menu close
back to Insight Page

Nanigans’ Dan Slagen on Predicting the Future

Written by: Cheryl Morris, VP Marketing

This article was originally published on Brand Innovators.

Today, most brand marketers optimize to CPA. Tomorrow, most brand marketers will optimize to ROI. The day after tomorrow, brand marketers will optimize to LTV.

It’s odd, because we used to talk about it all the time. It was the “next thing” and the “new way to think” about customer acquisition. Focusing on the lifetime value of a customer was our goal as performance marketers. As often happens with ideas that are before their time (or technology), however, lifetime value-based performance marketing fell by the wayside for most marketers as there just wasn’t a way to predict purchasing behavior and optimize at scale.

So instead, marketers settled. What did we settle for? CPM, CPC, CPL, CPA, ROI…whatever other KPI was being kicked around the annual conferences. The most efficient KPIs of the bunch were CPA and ROI, however, each had dangerous pitfalls for even the savvy marketers. The way to avoid those pitfalls is to focus on predictive lifetime value-based marketing. To see this, let me focus on the options available.

CPA: The concept of CPA is to acquire customers based on a set acquisition price. The inherent problem with CPA though is true value:

Consider two fabricated customers, C1 and C2 within the retail vertical.

C1: Single, male, 31, 0 children

C2: Married, female, 35, 2 children

C1 and C2 each see a t-shirt on sale for $10. The cost for an advertiser to acquire C1 is $2 while the CPA to acquire C2 is $4. Based on the nature of CPA modeling, C1 appears to be the more attractive customer and a CPA based optimization model will acquire more customers that look like C1 compared to C2 solely based on today’s cost. But is C1 a more valuable customer over time?

Return on Investment (ROI) Return on investment isn’t the wrong way to think about optimizing your performance marketing campaign, however, the way marketers think about ROI is shortsighted. Using C1 and C2 again, let’s consider a single time purchase today:

C1: Purchases a t-shirt for $14.99 and a sweatshirt for $29.99

C2: Purchases a t-shirt for $14.99

Based on today’s return, C1 again appears to be the more attractive customer. Hence, optimizing on single purchase ROI will again focus on acquiring customers that look and behave like C1 as opposed to C2. But again, is C1 a more valuable customer over time?

The good news for marketers is that we don’t have to optimize campaigns anymore to either CPA or ROI (and if you’re using CPM or CPC you should consider switching fields) and wonder if we’re truly acquiring the most valuable customers. The future is here now, so let’s go back to it and run our campaigns the way they should be optimized, via predictive lifetime value (LTV).

Predictive Lifetime Value (PLTV): The unfortunate truth for marketers that optimize to “today” rather than to “lifetime behavior” is that today’s behavior is not predictive of future behavior. One final analysis of C1 vs. C2 examines such future instances during a 4 week purchase cycle:


Today’s purchase: T-shirt ($14.99) & baseball cap ($19.99)

Purchases over the next 4 weeks: Zero additional purchases

Total purchase amount: $34.98


Today’s purchase: T-shirt ($14.99)

Purchases over the next 4 weeks: Sweatshirt ($29.99) & polo shirt ($49.99)

Total purchase amount: $94.97

As you can see, even though C1 appeared to be a better customer today and was less expensive to acquire, C2 was actually the more valuable customer with future purchases planned for her family. Marketers that can anticipate the difference between C1 and C2 using predictive lifetime value based optimizations will be the ones who build the most profitable customer base over time.

What’s the best part of predictive lifetime value based optimization? Besides being able to spend efficiently and scale with confidence, being able to report on expected revenue based on daily data tracking will ensure that any budget increases are solely based on positive ROI performance.

next post

But wait, there's more

Join Our Newsletter