How to Future-Proof Your Measurement

The privacy tipping point: Why future proofing measurement is no longer optional

Since the late 2010s, there has been growing recognition that user-tracking wouldn't remain a sustainable basis for marketing measurement forever.

However, a tipping point came in 2021 with Apple's iOS 14.5 release. This featured an update - App Tracking Transparency (ATT), requiring user opt-in to having their conversion journeys tracked.

With the vast majority of users opting-out, the Identifier for Advertisers (IDFA) - the primary data point for mobile marketing attribution - was immediately rendered useless at scale, and huge portions of the customer journey became invisible to measurement tools overnight.

For measurement approaches that rely on customer-journey information, like Multi-Touch Attribution and Last-Click variants, this has rendered them patchy at best. Recent estimates suggest they fail to properly measure 90% of your marketing activity, particularly upper-funnel awareness-building activities.

The impacts of this shift have been wide-reaching, and several primary trends are now emerging.

‍

What’s changing?

1. A decline in customer-journey based measurement

Measurement approaches relying on customer-journey information are increasingly subject to bottom-of-funnel bias, leaving them unreliable for full-funnel measurement.

Customer-journey touchpoints that occur close to conversion are less affected by the "information loss" from ATT and other privacy controls. As a result, attribution models based on customer journey analysis (like MTA, DDA, or Last-Click variants) are much more likely to "see" these touchpoints than top-of-funnel awareness-building interactions.

Since these models can only attribute to activity they can see, they significantly overestimate the value of bottom-of-funnel activity. As a result, estimates suggest they miss up to 90% of the true impact of upper-funnel marketing.

2. Finance teams are demanding causal certainty and aligned KPIs

The resulting instability in marketing data is amplifying frictions between Marketing and Finance teams.

Both teams historically appreciated customer-journey based measurement because of how fast they are to report and how little subjective interpretation was required in their attributions. For marketers, this meant they could support fast-moving digital optimizations; and for finance teams, the analysis of directly observed user-journeys brought auditable clarity and a sense of control.

With these methods no longer fully reliable, new ways of working are emerging from the alternatives: an increasing emphasis on the importance of causal analysis (proving that marketing activity actually caused changes to business outcomes and didn't just take place at the same time) and an alignment around KPIs with shared value to both CFOs and CMOs, like Incremental Profit and Growth Headroom.

‍

Traditional marketing metrics like ROAS (Return on Advertising Spend) and CPA (Cost Per Action) measure performance in isolation or retrospectively.

They work for tactical optimizations but lack the causal proof needed to justify investment decisions in financial reporting.

In contrast, Incremental Profit and Growth Headroom use causal measurement to forecast and prove financial outcomes—making them essential for boardroom conversations.

Incremental Profit — This is the true, additional profit generated solely by a specific marketing activity, after accounting for all associated costs.

Growth Headroom — This is the maximum capacity for profitable spending in a channel before hitting diminishing returns.

‍

3. Measurement is shifting toward approaches that are more causal and privacy-robust

To meet Finance teams' requirements and work within new privacy regulations, marketers are upgrading their measurement techniques.

This shift is moving marketing away from customer-journey based measurement toward more robust and causal options—specifically advanced Media Mix Modeling (MMM) and incrementality testing.

Media Mix Modeling — MMMs use statistical analysis of historical data to identify relationships between marketing activity and business outcomes. They're fundamentally privacy-robust because they work with aggregated data and don't track individual user journeys. Crucially, traditional MMMs are correlational: they show that marketing activity and sales happen together, but they can't definitively prove that one caused the other, making them prone to misinterpreting cause-and-effect.

Incrementality tests — These standalone experiments isolate the incremental impact of a single marketing variable, providing verifiable proof that Finance can trust. They compare two nearly identical groups: a Test group (exposed to your marketing) and a Control group (not exposed). Any observed differences between these groups can be confidently attributed to the marketing activity, proving causality.

While neither technique alone can form your complete measurement toolkit, together they're formidable: advanced MMMs now synthesize both methods, using the causal truth of incrementality tests to calibrate and refine the model's aggregated insights.

We explore MMMs, incrementality testing, and the advanced technique that synthesizes the two approaches in our previous lesson on building a suite of truth.

‍
4. AI and advanced statistics are enabling change

Traditional MMMs and causal incrementality tests are both historically slow to report. This latency makes them unsuitable for the agile, daily optimization needed for digital channels.

However, AI and more modern modeling techniques are overcoming this rigid speed barrier.

The core innovations are the adoption of Bayesian MMMs, a form of MMM based on a statistical approach called Bayesian modelling, and automations to speed up data ingestion and analysis.

In one of our previous lessons, we did a deep dive into daily and Bayesian MMMs and how they’re enabling change to marketing workflows.

The result of these changes is a measurement system built on privacy-robust data and causal methodology that is finally fast and flexible enough to meet CFOs' demand for certainty in an agile, changing market.

‍

The four pillars to future-proofing your measurement

These challenges are interconnected. Solving them requires a holistic framework built on four strategic pillars. Here are the practical steps to implement this future-proof approach:

1. Measurement data strategy

Move away from user tracking solutions (pixels, third-party cookies, IDFA) that face regulatory risk. Shift resources toward aggregated, privacy-first models like MMMs.

2. Methodological rigor

Make incrementality tests (like Geo-Lift) foundational. Use them not just as an audit, but as the mission-critical calibration input for your MMM.

3. Organizational alignment

Adopt shared financial KPIs based on causally proven incrementality metrics. This ensures every budget and strategic conversation with Finance is grounded in business value and causal certainty.

4. Technological agility

Choose modeling solutions that use advanced statistics (Bayesian) for direct, real-time calibration between privacy-robust MMMs and incrementality test results. Look for solutions that enable daily predictive forecasting through automated data ingestion.

‍

The bottom line

The key takeaway: user-journey based attribution is no longer reliable in a privacy-first world, forcing an industry-wide shift.

While technological solutions like Bayesian MMMs address the data gaps, the real solution is holistic. It requires not only technological upgrades but also an organizational shift in methodology and cross-functional alignment.