A Century-Old Problem That Still Plagues Marketing

John Wanamaker is today 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 criticism 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?

Two Root Causes Holding Companies Back

1. The Empowerment Gap: CMOs vs. Data Projects

NoMany CMOs face a critical disconnect: they’re responsible for revenue outcomes but lack authority over the technical infrastructure needed to measure them.

In a familiar scenario to CMOs, a  request for better attribution data from IT brings the following response: “That’s a six-month project requiring data engineers, new infrastructure, and six-figure investment.” The request gets deprioritized against other projects that are deemed to be more critical.

The result is that marketing teams default to what they can measure easily such as clicks, impressions, engagement while the CEO and board want revenue attribution. This disconnect creates what I call the Attribution Confidence Gap. This is the widening chasm between what marketing thinks is working and what’s actually driving revenue.

2. The Integration Nightmare: Too Many Systems, Too Little Connection

Modern marketing stacks resemble a technology hodgepodge rather than a unified ecosystem. A typical mid-sized company’s landscape looks something like this:

• 6+ ad platforms (Meta, Google, TikTok, LinkedIn, Reddit, etc.)
• 3+ sales systems (CRM, e-commerce, POS, partner portals)
• Multiple data formats (APIs, CSV exports, manual uploads)
• Dozens of campaign structures with inconsistent naming conventions

Each system speaks a different language, refreshes on different schedules, and guards its data with varying levels of accessibility. The technical debt from integrating these systems makes attribution projects feel like mining for rare earth elements, digging through layers of incompatible systems hoping to find coherent patterns.

Anatomy of a Modern Attribution Solution

Part 1: The Data Aggregation Layer

This is where tools like Supermetrics become invaluable. (We became a Supermetrics partner this year.)They solve the first-mile problem by collecting marketing data from dozens of platforms into a single location. Think of Supermetrics as the universal translator for your marketing data, it speaks Meta, Google, TikTok, and 50+ other platform languages, outputting everything in a consistent format.

Supmetrics solves a big headache. No more manual daily exports from 15 different platforms. No more spreadsheet consolidation nightmares. No more wondering if you’re comparing apples to apples across channels.

But aggregation is just the beginning. Raw data in one place doesn’t equal insights.

Part 2: Data Engineering

This is where data engineers enter the picture to put it all together. Once data is aggregated, it needs:

1. Transformation & Cleaning

• Standardizing naming conventions across campaigns
• Handling timezone conversions
• Dealing with API changes and platform updates
• Managing data gaps and inconsistencies

2. Business Logic Implementation

• Defining what constitutes a “conversion” in your business context
• Establishing attribution windows (1-day click? 7-day view? 28-day assisted?)
• Creating hierarchical rules for multi-touch scenarios
• Building in seasonality and external factor adjustments

3. Data Modeling for Analysis

• Designing star/snowflake schemas optimized for marketing queries
• Creating slowly changing dimensions for campaign metadata
• Building aggregate tables for performance
• Implementing data quality monitoring and alerts

4. Visualization & Democratization

• Building role-based dashboards (CMO vs. media buyer vs. finance)
• Creating self-service exploration capabilities
• Setting up automated reporting and anomaly detection
• Ensuring data governance and access controls

Critical Decision Points: What CMOs Need to Know

Before embarking on an attribution journey, CMOs must prepare for several key decisions:

Decision 1: Causation vs. Correlation

Causation (deterministic attribution): “This $1 Facebook ad directly led to this $50 sale.” Correlation (probabilistic attribution): “When we increase Facebook spend by X%, sales increase by Y% with Z% confidence.”

Most businesses need to accept correlation-based attribution due to:

• Platform restrictions (no pixel placement on partner sites)
• Offline conversions
• Complex customer journeys
• Data privacy regulations

In our conversations with clients, they do understand that they can start with correlation, be transparent about confidence levels, and work toward causation where possible.

Decision 2: Solving ROAS Attribution vs. Waiting for IT

CMOs have a leadership decision to make. Either they can accept that ownership of solving the attribution challenge sits within the marketing domain, using tools and partners that don’t require extensive IT involvement. Or they can wait for prioritization against other IT projects, and hope marketing attribution makes the cut and that IT has the data engineering resources to tackle the problem.

The IT queue reality is that IT priorities naturally skew toward security, infrastructure, and revenue-critical systems. Marketing attribution often gets classified as “nice to have” analytics.

But waiting to get prioritized has a huge impact on the company. Waiting means not only inefficient marketing spending, but also opportunity cost. Sales growth that marketing could be enabling is lost.

What CMOs have to accept is that attribution is a revenue optimization project. Just as marketing doesn’t wait for IT to choose advertising platforms or design campaigns, marketing shouldn’t wait for IT to measure campaign effectiveness. The tools and partners exist today to enable marketing-led attribution.

Decision 3: Accuracy vs. Actionability

Perfect attribution is a myth. In the real world, systems are layered, data is messy, and customer journeys are complex. The pursuit of 100% accuracy often becomes the enemy of good enough to act.

The real goal is actionable accuracy where data is reliable enough to change decisions with confidence, even if it contains some uncertainty.

One of our clients is highly dependent on social advertising. They face a constant challenge because platform algorithm changes can fundamentally shift quarterly revenue economics overnight.

With actionable accuracy (not perfect attribution), they can see the ROAS drop in real time and use A/B testing to shift their spending on ads that are working. Without it, they would have wasted weeks waiting for more data while burning through budget on what had become an inefficient ad.

The Marketing Mandate

If you’re still in the era of spray and pray marketing, you now have smarter options. Not because we have perfect attribution, but because we finally have enough attribution to make better decisions.

The companies winning today don’t have perfect data. They’re the ones who have moved from “we don’t know” to “we know enough to act.” They’ve bridged the gap between marketing execution and revenue measurement, not with perfect solutions, but 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, enabling marketers to waste less each quarter while scaling what works.

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.

Ren Agarwal is CEO at San Francisco based StoryAZ Studio and Chipsy.io