The Next Dollar
What marginal economics can teach us about marketing measurement.
The real value of marketing measurement isn't assigning credit. It's improving capital allocation. For years, marketing measurement conversations have been dominated by attribution.
Which channel deserves credit?
Which touchpoint influenced the conversion?
Which attribution model is most accurate?
These are important questions. Attribution provides a framework for understanding how outcomes are distributed across marketing activities. But attribution was never the destination. The ultimate purpose of measurement is not to assign credit. It is to improve decision quality.
The reason I care about attribution, incrementality, and increasingly multi-touch attribution (MTA), has very little to do with redistributing credit across channels. My interest is much more practical:
Can measurement help us make better investment decisions?
Because in the end, growth is not created by reporting. Growth is created by allocating capital more effectively than we did yesterday.
The Original Question
Earlier in my career, one of the most important questions I faced wasn't an attribution question at all. It was an investment question.
How much can we profitably invest in customer acquisition before the economics stop creating value?
To answer it, we built an economic model that incorporated acquisition costs, customer lifetime value, and net present value. The objective wasn't measurement for measurement's sake. The objective was determining where capital should flow. Once we identified the NPV breakeven point, we had a rational framework for setting investment levels. At the time, the measurement system was doing exactly what it needed to do.
It helped us make a better decision.
New Questions Emerge
As measurement capabilities mature, organizations often begin uncovering insights that their original frameworks cannot fully explain. In our case, some analyses suggested that non-brand search wasn't only driving customers who converted directly through search. It also appeared to be influencing behavior that surfaced later through other channels.
For example, some customers who initially engaged through non-brand search also responded to direct mail. That raised an interesting question.
If search is creating value that ultimately materializes elsewhere, are we underestimating the value created by search? The conversation quickly turns into an attribution question. Should non-brand search receive 100% credit? 120% credit? Something else entirely?At first glance, that seems like the right question. But I don't think it is.
Attribution Assigns Credit. Investment Decisions Allocate Capital.
The more I have encountered these discussions throughout my career, the more I've realized we are rarely debating attribution. We are debating economics.
The question isn't:
How much credit should search receive?
The question is:
Are we underestimating the economic value created by search?
Those sound similar, but they lead organizations down very different paths. Attribution is fundamentally a credit assignment exercise. Investment decisions are capital allocation exercises. One helps explain what happened. The other determines what happens next.
Conflating the two often creates more confusion than clarity.
Enter Marginal Economics
This is where the concept of marginal value becomes useful.
One of the foundational ideas in economics is that decisions should be made based on the value of the next unit, not the average value of all units. Businesses hire the next employee. Invest in the next project. Spend the next marketing dollar. Everything happens at the margin.
Most marketing discussions focus on averages.
Average CPA.
Average ROAS.
Average conversion rate.
Average attribution.
Marginal economics asks a different question:
What value will the next dollar create?
That subtle shift changes how we think about measurement. The purpose of measurement is not to perfectly explain the past.
The purpose of measurement is to improve future decisions.
The Real Value of MTA
This realization changed how I think about multi-touch attribution. For years, MTA has been marketed as a way to produce more accurate attribution. While that may be true, I think that framing misses the bigger opportunity. The biggest opportunity in MTA is not attribution accuracy.
It is decision accuracy.
Attribution tells us how value appears to be distributed. MTA has the potential to reveal value creation that traditional measurement frameworks fail to recognize. When that happens, the benefit isn't a prettier attribution report. The benefit is a better investment decision. If a channel creates downstream demand that traditional attribution fails to recognize, then the value of that channel may be understated. Not because it deserves more credit. But because our economic model may be underestimating the value created by the next dollar invested.
That's a fundamentally different objective. The goal isn't to determine which channel wins the attribution argument.
The goal is to determine where capital should flow.
The Evolution of Measurement
Looking back, I've realized that every meaningful step in measurement maturity follows the same pattern. Attribution helps us understand what happened. Incrementality helps us understand what would not have happened. MTA helps us understand how value is distributed across touchpoints. Economic models help us understand profitability. Marginal analysis helps us understand where the next dollar should go. Each stage provides a more complete picture than the one before it. But none of them are the destination. The destination is better decision-making.
A Better Framework
Many measurement discussions eventually become debates about credit allocation. Those conversations have value. Organizations need consistent frameworks for understanding performance and evaluating marketing activity. But attribution alone does not create growth. Growth comes from making better investment decisions. That's why I view attribution, incrementality testing, MTA, and economic modeling as complementary rather than competing disciplines.
Each answers a different question.
• Attribution helps explain outcomes.
• Incrementality helps establish causality.
• MTA helps uncover how value flows across touchpoints.
• Economic models help quantify profitability.
• Marginal analysis helps determine where future investment should go.
The organizations that create sustainable growth are rarely the ones with the most sophisticated attribution model. They're the ones that build measurement systems capable of improving capital allocation decisions over time. Because the ultimate goal of measurement is not to win the attribution debate.
The goal is to make the next investment decision better than the last one.