Upper-funnel measurement uses incrementality testing to show what top-of-funnel spend contributes, not just what last-click reporting credits. This matters because brands that lean heavily on last-click data undervalue upper-funnel media and cut spend that was working. Without incrementality data, platforms can exploit vague goals like video views, delivering cheap traffic that never converts. This video explains how to measure upper-funnel media on a performance basis and set fitting ROAS targets.
Why Upper-Funnel Measurement Needs Incrementality Testing, Not Last-Click Data
Upper-funnel measurement needs incrementality testing, not last-click data, because brands leaning on last-click attribution routinely overweight the bottom of the funnel and end up cutting upper-funnel spend that was quietly working. In this conversation, LiftLab’s Jon Lorenzini and John Wallace talk with growth marketing entrepreneur Adam Lovallo about the tension between what Lorenzini calls lean-back media and lean-in media. Wallace explains that most brands reaching out to LiftLab already suspect, deep down, that they’re overweighting last-click attribution, and the job becomes showing them the performance-based incrementality of their middle and upper-funnel spend before they cut it.
Lovallo adds a pointed observation from his own client conversations: when a campaign is optimized for video views instead of conversions, CPMs drop dramatically and completion rates spike, but click-through rates collapse to nearly zero, because platforms are extremely good at finding an audience that will watch and never click. His prescription is blunt: if you’re going to invest more in top-of-funnel, lower the ROAS target you’re holding it to rather than judging it by bottom-funnel standards, or platforms will happily spend your budget on their worst-converting audience. Wallace’s response is that the fix isn’t abandoning the idea, it’s measuring the actual incremental revenue each version of that spend produces.
In this video, you’ll learn:
Why over-reliance on last-click attribution harms long-term growth
How to assess the performance-based incrementality of upper-funnel media
The dangers of optimizing campaigns purely for video views or impressions
Why platforms often show ads to non-converting audiences when misaligned
Strategies to adjust ROAS targets to protect top-of-funnel investments
TIME STAMPFull Video Transcript: Upper-Funnel Measurement and Incrementality Testing Explained
+–
In this video, LiftLab’s Jon Lorenzini and John Wallace talk with growth marketing entrepreneur Adam Lovallo about why upper-funnel media needs to be measured on incrementality, not last-click attribution.
[0:00 – 0:17] The Balance of Branding and Performance
[Jon Lorenzini]: I know we’ve run quite a bit of branding tests, performance tests, and I think you’re right where it’s a little mixture of both. I call it lean back media versus lean-in media. So, John, what are your thoughts about this allocation, because I think this is the number one question you deal with with our clients here at LiftLab.
[0:17 – 1:36] Overcoming Last-Click Attribution
[John Wallace]: My answer is that quite often the brands that are reaching out are so over dependent on looking at click-based, last-click based attribution, that they know in their hearts that they are probably overweighting that end of the funnel. So, our “let go of the drug,” kind of pull it, rip it out of their hands, is to show them the performance-based incrementality really of their middle and upper funnel medium, and we kind of explain, look, it may be driving some other things that we’re not capturing here. It may have some even longer-term effects. You may want to supplement this with some survey-based data, have at it, but at a minimum, if you cut this off, this is how much revenue you’re going to see go poof. And then it gets back to what Adam kind of said, you know, then you get to sit and look at it kind of objectively and say, now are those extra dollars into top of funnel more effective than what I’ve already kind of done, or my target for what you’d call maxing out one of the social channels or our search? Can we measure upper funnel on a performance basis, not to say that’s all it delivered, but to say, you know, you may be shooting yourself in the foot if you cut this off. If you just kind of walk in and say cut spend, and people do it top of funnel, then they need to know what’s going to be the outcome of that.
[1:37 – 3:25] The Mechanics of Top-of-Funnel Advertising
[Adam Lovallo]: I want to add something, because one thing I found is, often when I have this conversation, which I’ve had a million times with CMOs, they’re like, okay, we’re gonna invest more in top of funnel. I’m like, great, that’s awesome, I think the bottom of the funnel is the most [ __ ], so I’m with you, we’re gonna do more brand. So some amount of the campaign is going to optimize for video views, some amount of the budget is going to optimize for traffic, and then some amount of budget is going to optimize for conversion. And I’m like, okay, we can do that for sure, then we can measure it in the way you just described, done, and that’s cool, and it could work out great. But isn’t it notable that when we optimize for video views, for example, the CPMs of that media are a tenth of the media that we’re optimizing for conversions? And also, isn’t it remarkable that the video view completion rate for those campaigns is sky high relative to our normal campaigns, and the click-through is almost zero? It’s almost like the algorithms are so good at showing ads to people who just watch stuff but never click, that when advertisers ask for people who just watch stuff and never click, it’s like, okay, let’s move up the chain, you want to optimize for just impressions? Great, wow, now the video view rate is almost zero percent on the exact same creative, click-through rate zero, and the CPMs are even lower. So that’s where I tend to be like, okay, you want to put more into the top funnel, spectacular, let’s lower our ROAS targets for our actual campaigns that drive results, and instead of targeting a 1.2x one-day click, whatever, you target a 0.5x, because if you give some of these platforms the opportunity to take your money and show it to their worst audience, they will jump at the opportunity.
[3:25 – 4:02] Measuring Incremental Revenue
[John Wallace]: They will, but let’s know the incremental revenue for those. I mean, I love that you’re bringing up the options, because when we say something like top of funnel, there are still media buying choices there, not as many, but your examples are great. Sure, I would just say, you know, I want to know the incremental of each of those, right. I mean, absolutely fine with me that you got zero people clicking, but you’ve got a healthy chunk of something else happening.
[Adam Lovallo]: If you can show that a video view campaign on Facebook is actually driving ROAS, first of all, CMOs will be stoked, you guys look like geniuses, and you know what, the people at Facebook, they’ll be like, whoa, this is amazing, everybody should do this.
Key Lessons: Measuring Upper-Funnel Media With Incrementality Testing
Attribution strategy: Move beyond last-click models to understand the true incremental value of upper-funnel efforts.
Incrementality testing: Use testing to prove revenue impact and prevent unnecessary budget cuts for brand-focused campaigns.
Algorithm optimization: Be cautious with vanity metrics like video completion rates, which often result in low-quality traffic.
Platform incentives: Recognize that advertising platforms will exploit misaligned objectives to prioritize cheap, non-converting impressions.
Performance goals: Adjust ROAS targets for top-of-funnel campaigns to reflect their unique role in the customer journey.
