John Wanamaker is remembered for his observation: "Half the money I spend on advertising is wasted; the trouble is I don't know which half." You would think this should be a historical artifact by now. Yet, ask any CMO today, and you'll hear a strikingly familiar frustration: "I know my marketing is driving sales, I just don't know how to show it."
Why, as we approach 2026, does marketing attribution remain one of the most persistent and expensive problems in business?
Many CMOs face a critical disconnect: they're responsible for revenue outcomes but lack authority over the technical infrastructure needed to measure them. A request for better attribution data from IT often returns: "That's a six-month project requiring data engineers, new infrastructure, and six-figure investment."
The result is that marketing teams default to what they can measure easily — clicks, impressions, engagement — while the CEO and board want revenue attribution. This creates the Attribution Confidence Gap: the widening chasm between what marketing thinks is working and what's actually driving revenue.
Modern marketing stacks resemble a technology hodgepodge rather than a unified ecosystem. A typical mid-sized company's landscape includes:
Each system speaks a different language, refreshes on different schedules, and guards its data with varying accessibility. The technical debt from integrating these systems makes attribution projects feel like mining for rare earth elements.
This is where tools like Supermetrics become invaluable. They solve the first-mile problem by collecting marketing data from dozens of platforms into a single location — the universal translator for your marketing data, speaking Meta, Google, TikTok, and 50+ other platform languages in a consistent format.
Once data is aggregated, it needs transformation and cleaning (standardizing naming conventions, handling timezone conversions, managing data gaps), business logic implementation (defining conversions, establishing attribution windows, building multi-touch rules), and data modeling for analysis (designing schemas, building aggregate tables, implementing data quality monitoring).
Causation: "This $1 Facebook ad directly led to this $50 sale." Correlation: "When we increase Facebook spend by X%, sales increase by Y% with Z% confidence." Most businesses must accept correlation-based attribution due to platform restrictions, offline conversions, complex customer journeys, and data privacy regulations.
CMOs have a leadership decision to make: accept ownership of the attribution challenge within the marketing domain using tools that don't require extensive IT involvement, or wait for IT prioritization. The IT queue reality is that marketing attribution often gets classified as "nice to have." Waiting means not only inefficient marketing spending, but lost sales growth opportunity.
Perfect attribution is a myth. The real goal is actionable accuracy — data reliable enough to change decisions with confidence, even if it contains some uncertainty. One client highly dependent on social advertising can now see ROAS drops in real time and shift spending to working ads. Without it, they would have burned budget for weeks.
The companies winning today don't have perfect data. They've moved from "we don't know" to "we know enough to act." They've bridged the gap between marketing execution and revenue measurement with practical, phased approaches that deliver increasing clarity over time.
Wanamaker's problem isn't solved by finding which half is wasted — it's solved by continuously reducing the unknown portion through better measurement. The question for 2026 isn't "Can we achieve perfect attribution?" It's "How much better could our decisions be with 70% more attribution clarity?" For most companies, that 70% is worth millions.