Last-click attribution is reliably wrong because it only counts the final touchpoint, ignoring how consumers interact with ads. This happens because ad platforms already run their own optimization algorithms, so marketers need an independent scoring model. Without that check, budgets stay locked into whichever platform reports the most credit, even as efficiency declines. This video explains how to build a scorekeeper model that holds ad platforms accountable for future performance.
Why Last-Click Attribution Is Wrong for Measuring Ad Performance
Last-click attribution is reliably wrong for measuring ad performance because it counts only the final touchpoint in a nonlinear consumer journey, missing everything that led up to it. In this video, John Wallace, CEO of LiftLab, equates a counting algorithm to a knife in a gunfight. He believes that being consistently wrong in a predictable way is not the same as being trustworthy, and that marketers need to graduate from counting-based attribution to a model-based, nonlinear approach. Just as ad platforms already run their own optimization algorithms on a marketer’s behalf, marketers need an independent scorekeeper algorithm to monitor efficiency across every channel they use.
The goal is to proactively shift budgets by rewarding platforms that are performing efficiently and pulling back on the ones that are not, rather than defaulting to whatever an ad platform’s own algorithm recommends. Since no ad platform is going to call and volunteer that its performance has slipped, Wallace argues that holding platforms accountable for what they can deliver next, not just their historical averages, falls entirely on the marketer.
In this video, you’ll learn:
Why the last-click attribution model is considered reliably wrong
The necessity of adopting model-based, nonlinear algorithms for ad performance monitoring
How to hold ad platforms accountable by optimizing spend based on future potential rather than historical averages
The importance of manual oversight in managing automated budget-allocation algorithms
TIME STAMPFull Video Transcript: Last-Click Attribution Explained
+–
In this video, LiftLab CEO John Wallace breaks down why last-click attribution fails marketers, and what a model-based scorekeeper approach looks like in practice.
[0:00 – 0:17] The Problem with Last-Click Attribution
The reason last-click is insidious, like it’s everywhere, is that it’s a counting approach. We can count up how many things that happen. It’s reliably wrong or it’s consistently wrong, and there’s a certain confidence that comes with being reliably wrong. I challenge that.
[0:17 – 0:47] Moving to a Model-Based Approach
To graduate from a counting algorithm we need something that is model-based, something that’s nonlinear. You’re already doing this with all of your ad platforms. You’re giving your money into PMax campaigns or Advantage Plus campaigns and letting the algorithm do the work for you.
What you need is one algorithm that’s kind of the scorekeeper for all of these other algorithms. Without an algorithm to monitor the algorithms you’re spending on, you’re little equipped; you’re kind of bringing a knife to a gunfight.
[0:47 – 1:10] Accountability and Optimization
Embracing these concepts that I need to hold all of the ad platforms accountable not just for their average performance, but what they can do for me next week if we increase our spend. Now we’re in a position to, as I like to say, take from the poor and give to the rich. We want to reward those platforms where we’re too efficient, and we want to dial back the platforms where we’ve become inefficient.
[1:10 – 1:49] The Marketer’s Responsibility
Think about it, you’re already doing this. You’re giving your budget to algorithms. It’s just that those algorithms are operated by your ad platform partners, and you need to keep doing that; there’s no issue with it.
You need to be equipped in some shape or form to monitor that performance. They’re never going to ring you up and say, “Thank you for spending a million dollars last month; you should have spent $900,000, and the last $100,000 were inefficient.” They’re not going to call you up and say that. It falls on you as a marketer to have some method to know when to say when and how to reward the platforms that are still what I call too efficient, that there’s room to grow or scale those platforms.
Key Lessons: Building a Scorekeeper Model for Ad Spend Optimization
Move beyond counting: Don’t rely solely on simple counting metrics; use advanced modeling to understand the true impact of your marketing spend.
The scorekeeper approach: Implement an independent algorithm to monitor and evaluate the performance of the various ad platform algorithms you use.
Take from the poor, give to the rich: Actively reallocate budgets from inefficient channels to those that show room for scalable, efficient growth.
Marketer accountability: Ad platforms will not tell you when you are overspending; it is your responsibility to monitor performance and adjust accordingly.
