Part 4: Operations · Chapter 15

Attribution Reality

9 min read

Core Argument: You cannot measure what you think you are measuring. Attribution is fundamentally broken, and the industry pretends otherwise. Accepting attribution uncertainty is the first step toward practical measurement.

The Attribution Illusion

The marketing technology industry has sold a comfortable lie: that we can precisely attribute revenue to specific marketing touchpoints.

This narrative is fiction. The reality is complex. Buyers see ads but do not click. They hear mentions on podcasts. They ask colleagues. Multiple touchpoints influence, but attribution models assign credit based on arbitrary rules, not causal influence.


Why Attribution Is Broken

Problem 1: Multi-Touch Reality

B2B purchases involve 6-10 touchpoints on average. Attribution models assign 100% credit to one touchpoint or distribute credit using arbitrary formulas. The formulas are made up.

Problem 2: The Dark Funnel

40-60% of B2B influence happens outside trackable touchpoints: podcasts, word of mouth, community discussions, social browsing. The dark funnel is invisible to attribution.

Problem 3: Privacy Degradation

iOS 14.5, cookie deprecation, and privacy regulations have reduced tracking capabilities by 30-50%. Attribution models built on declining data become less accurate over time.

Problem 4: Platform Self-Reporting

Every platform over-claims. Add up platform-reported conversions, and you will exceed actual conversions by 50-200%.


Practical Attribution Approaches

Approach 1: Directional, Not Precise

Use attribution for directional guidance, not precise allocation. Accept "LinkedIn appears to influence pipeline" but reject "LinkedIn drove exactly $347,421."

Approach 2: Multiple Models

Run multiple attribution models simultaneously. Convergence across models indicates signal. Divergence indicates noise.

Approach 3: Incrementality Testing

Test whether a channel actually causes conversions through holdout experiments. Divide audience into test and control, compare conversion rates, calculate incremental lift.

Approach 4: Self-Reported Attribution

Ask buyers: "How did you hear about us?" This captures dark funnel influence that tracking misses. The simplest method is often the best.


Building a Measurement Framework

Layer 1: Platform Reporting - Use for relative comparisons within platforms.

Layer 2: CRM Attribution - Track leads through CRM with source attribution. Your primary measurement.

Layer 3: Self-Reported Attribution - Capture buyer-reported influence for dark funnel visibility.

Layer 4: Incrementality Testing - Periodically validate major channels for incremental impact.

Layer 5: Aggregate Analysis - DER (Demand Efficiency Ratio) is the ultimate metric. Individual channel attribution may be uncertain; total efficiency is measurable.


Conclusion: Embrace Uncertainty

Attribution is broken. It will not be fixed. Privacy trends are making it worse. The response is not to give up on measurement. It is to embrace uncertainty and use multiple imperfect signals.

Precision is an illusion. Directional accuracy is achievable. Optimize for actionable, not perfect.

Key Frameworks

The Dark Funnel
All buyer influence that happens outside trackable touchpoints: podcasts, word of mouth, social browsing, community discussions. Estimated at 40-60% of B2B influence.
Multi-Layer Measurement
The framework combining platform reporting, CRM attribution, self-reported attribution, incrementality testing, and aggregate analysis.
Incrementality Testing
Holdout experiments that test whether a channel causes conversions by comparing test and control groups.

References

  1. Gartner (2024). Marketing Attribution in a Privacy-First World. Link
  2. Forrester (2023). The State of B2B Attribution. Link