Marketing KPI measurement works when a metric responds to marketing levers with a causal link to sales. This matters because a metric that fluctuates but adds no value, what Pauwels calls chasing squirrels, pulls budget from revenue drivers. Without a mental model of how a business functions, teams default to benchmarks that miss their brand size. This video explains how blending modeling with experiments produces KPIs a business can trust.
Why Marketing KPI Measurement Depends on Causality, Not Just Movement
Marketing KPI measurement depends on causality, not just movement, because a metric can swing wildly without ever actually explaining what’s driving sales. In this Curve Your Enthusiasm conversation, Jon Lorenzini and John Wallace host Dr. Koen Pauwels of Northeastern University, editor-in-chief of multiple marketing journals, for a session built around bridging academic rigor with day-to-day marketing practice.
Pauwels opens by dismantling reliance on industry-wide benchmarks: a finding like large brands seeing little incremental value from paid search, true for a company like eBay with existing brand awareness, simply doesn’t generalize to a smaller or niche brand where search still matters enormously. He recommends companies run their own field experiments rather than borrow someone else’s average. Working through case studies on office furniture and cosmetics, he introduces his “MEM,” Model-Experiment-Model, approach: models generate a hypothesis, experiments test it in the real world, and the results feed back into a sharper model, which is also how he approaches shifting spend away from inefficient channels like Direct Mail once customer segmentation, new versus existing, reveals where it’s actually working. On KPIs specifically, Pauwels sets a clear bar: a valid metric has to respond to marketing levers and hold a real causal relationship with sales, warning against what he calls “chasing squirrels,” metrics that move constantly but carry no real business signal. He closes by reminding the audience that perfect measurement is far less important than making informed decisions from whatever blend of modeling and experimentation is actually available.
In this video, you’ll learn:
How to avoid the pitfalls of relying on industry-wide marketing benchmarks
Why experimental field studies are crucial for testing marketing incrementality
Techniques for optimizing media spend across offline and digital channels
How to differentiate between vanity metrics and true performance KPIs
The importance of building an iterative, decision-oriented marketing mental model
TIME STAMPVideo Overview: Marketing KPI Measurement Summarized
+–
[0:00 – 2:12] Introduction
Jon Lorenzini introduces the Curve Your Enthusiasm series and welcomes special guest Dr. Koen Pauwels, a Distinguished Professor at Northeastern University and editor-in-chief of multiple marketing journals. The episode aims to bridge academic rigor with marketing practice.
[2:13 – 10:45] Paid Search Incrementality and Brand Size
Dr. Pauwels discusses the nuance of paid search effectiveness. He highlights that while some large brands, like eBay, found paid search had little impact due to existing brand awareness, this does not apply to smaller companies or niche products where search is vital. He emphasizes that companies should conduct their own field experiments rather than relying on industry-wide averages.
[10:46 – 28:09] Customer Touch Points and Media Mix
This segment focuses on two case studies: InEffect (office furniture) and L’Occitane (perfume and cosmetics). Dr. Pauwels explains the “MEM” (Model-Experiment-Model) approach. By moving away from inefficient Direct Mail and optimizing media spend based on customer segmentation (new versus existing), companies can achieve better ROI and growth. They also discuss the danger of over-focusing on CAC (Customer Acquisition Cost) versus LTV (Lifetime Value).
[28:10 – 42:06] Consumer Attitude Metrics and KPIs
Dr. Pauwels outlines criteria for valid KPIs. He argues that metrics must be responsive to marketing levers and have a causal relationship with sales. He warns against “chasing squirrels,” metrics that fluctuate rapidly but hold no real business value, and advocates for building a clear mental model of how the specific industry functions.
[42:07 – 55:20] Rapid Fire Q&A and Closing
Dr. Pauwels addresses questions regarding reach versus frequency, the persistence of TV as a marketing lever, and the ongoing challenge of attribution. He reiterates that perfection in measurement is secondary to making informed decisions through a blend of modeling and experimentation. The session concludes with a look ahead to future forum dates.
Key Lessons: Choosing Marketing KPIs That Actually Explain Sales
Measurement strategy: Use a model-experiment-model approach to validate performance rather than blindly trusting attribution software.
Brand impact: Recognize that established brands see different search effectiveness compared to smaller, growth-focused companies.
KPI selection: Prioritize metrics that demonstrate causal relationships with sales instead of tracking engagement data that changes too quickly.
Customer segmentation: Tailor communication channels based on whether you are targeting new prospects or re-engaging dormant existing customers.
Resource allocation: Implement a marginal return analysis to move budget away from saturated channels toward those with higher potential.
