pLTV: What are we trying to prove?
This article discusses how predictive Lifetime Value (pLTV) targeting can enhance advertising strategies by focusing on long-term customer value rather than immediate returns. It outlines an experimental approach where existing campaigns are split into two: one using traditional optimization methods and the other employing pLTV-based optimization, with the goal of demonstrating that pLTV targeting leads to significant improvements in long-term Return on Ad Spend (ROAS) and can be effectively scaled. The article also details methods for measuring the success of pLTV implementation and explores what integrating pLTV as a core component of user acquisition strategies entails.
pLTV-based targeting improves how ad networks identify valuable user groups. Traditional targeting focuses on “early value users” — people who produce revenue soon after seeing an ad. pLTV-based targeting expands this approach in two ways: it helps networks rank the difference between early value users and those who keep coming back. It also spots users who may not produce returns right away but are likely to become valuable over time.
While standard targeting focuses on early signs of value, pLTV adds a forward-looking signal into the network’s decision-making, balancing early wins with long-term growth.
pLTV success through experiments
To prove the positive impact pLTV can have on your growth, we will split one of your existing evergreen campaigns into two variants - both having the same targeting. One, acting as our baseline, will optimize towards your business as usual goal and the other towards Churney’s pLTV event. The experiment will focus on two different business goals.
Goal 1: Long-term ROAS improvements to an existing campaign
During the initial phase, we focus on demonstrating that pLTV-based optimization works and leads to significant long-term ROAS (return on ad spend) gains. We aim to maintain similar spending levels between the two test arms to ensure a fair comparison of results. Typically, businesses see an initial minor ROAS drop in the early days post-acquisition compared to usual targeting methods, followed by a steady improvement & overtake in ROAS of the pLTV campaign over the baseline. For us, anything below a double-digit increase in long-term ROAS compared to your business-as-usual targeting means we failed to deliver on our promise.
Goal 2: Prove success can be scaled
Experience shows that once customers see ROAS improvements from pLTV, many feel tempted to lower ROAS targets or ramp up spending to scale quickly. However, we recommend holding off on these rapid changes during the experiment.
For the scaling part of the experiment, we gradually increase the spending on the Churney pLTV campaign while maintaining a baseline spending on your existing business-as-usual (BAU) one. We define success as the pLTV campaign continues to outperform your BAU campaign in terms of ROAS even as the spend on pLTV is increased to match or exceed your previous total spend when only running your BAU campaign.
Our agenda is to maintain some percentage of spend on your BAU campaign to serve as a reference point to compare pLTV performance against. By gradually reallocating more spend to pLTV while leaving some BAU spend, we can evaluate if pLTV’s relative performance advantage holds as its spend scales up.
Measuring experiments
Churney considers attribution outside of the pLTV scope and separate from pLTV modeling. Our focus is predicting lifetime value using your zero and first-party data without getting involved in attribution. We are adjacent to attribution models - whatever model you employ, we accept as truth. With that in mind, there 3 main sources from which pLTV performance can be measured:
Ad platform analytics - looking at the short-term metrics like cost per install, and cost per purchase, and evaluating the pLTV numbers as attributed by the ad platform.
Ad platform experiment framework - major ad networks offer frameworks that support experiments. In Meta, we conduct a conversion lift study to measure the incremental lift driven by pLTV. In Google, we use the Google Ads Experiments framework.
Your DWH or MMP - for the subset of users that can be attributed to specific campaigns based on clicks, we will follow up on their actual long-term behavior and value from your data warehouse or mobile measurement partner (MMP) data.
Success! pLTV as part of your ongoing UA strategy
Once we’ve shown that pLTV can significantly boost your long-term ROAS and translate into overall profit growth, we wrap up the experiment phase and integrate pLTV as a core part of your targeting strategy.
Our ultimate goal in supporting you over the long term is to find the ideal balance between two approaches: your current business-as-usual targeting, which focuses on early, predictable values, and Churney’s pLTV targeting, which emphasizes long-term potential. This optimal mix could range from fully leaning on Churney’s approach, spending all or most of your budgets on pLTV campaigns, to a combination where business-as-usual targeting still plays a larger role. By blending both strategies, we aim to maximize your profits by harnessing the unique strengths of each approach.