Traditional Multi-Touch Attribution (MTA) models are no longer sufficient due to significant changes in the privacy landscape. For example, the introduction of the CCPA, GDPR, and recent iOS 17 updates that block tag managers and remove tracking parameters for users browsing in privacy mode on their iOS devices.

These changes have reduced the granularity of customer tracking and online data collection, compelling advertisers to rethink their ad targeting and attribution methods. Consequently, there is a growing need for new measurement tools that can accurately quantify the true impact of marketing campaigns (incrementality).

One of the most effective being…

Matched Market Testing (MMT)

Matched market tests have emerged as a vital solution in this context. They offer the closest available option to replicate real lab-based experiments in a real-world setting. Unlike slow Marketing Mix Models (MMMs), a matched market test can provide a speedier solution, allowing marketers to learn fast, calibrate their MMMs, and validate MTA modeling. This is crucial, as the speed of obtaining actionable insights can significantly impact the effectiveness of marketing strategies.

Advertisers are now rethinking their methods, and incrementality studies have become essential. However, not all incrementality studies on different platforms are equal, as they offer varying tools and methodologies. This variation necessitates a cohesive tool that ensures comparability across different platforms and campaigns.

In the following part of this article, we will dive into the current options, why marketers need a new measurement approach, and which model to choose (and when).

Why Ad Platforms’ Lift Tests Are Not Always Suited

While ad platforms offer “lift tests” (incrementality tests) to measure the impact of campaigns, these tests are not always suited for all marketing needs. Here are several reasons why:

Inconsistent Availability and Methods Across Platforms

Ad platforms often have different methodologies and levels of availability for incrementality testing. For example, Meta might use a different approach to Google when it comes to setting up control groups and measuring lift. This inconsistency can make it challenging for marketers to compare results across different platforms accurately.

Difficulty Reconciling Platform-Specific Incrementality with First-Party Data

Reconciling platform-specific incrementality results with first-party (1P) data can be problematic. Each platform has its own way of attributing conversions, which might not align with a company’s internal metrics. For instance, you might find discrepancies between the sales lift reported by a platform and their actual sales data, leading to confusion and potential misallocation of your marketing budget.

High Budget Requirements for Statistical Significance

Achieving statistical significance in platform incrementality tests often requires a substantial budget. Small to medium-sized businesses might not have the financial resources to run these tests effectively.

Complex Marketing Mixes (Difficulty Managing Incrementality Programs Objectively)

Modern marketing strategies often involve a complex mix of channels and tactics. Managing incrementality programs across these diverse elements can be challenging. For instance, a cross-channel campaign that spans Social Media, Search, and Display advertising might produce results that are difficult to disentangle when using platform-specific incrementality tests.

Limits for Cross-Platform Incrementality Measurement

Ad platform incrementality tests are typically limited to measuring the impact within their own ecosystems. This limitation makes it hard to assess the cross-platform incrementality. For example, if a company runs campaigns on both Meta and Google, it’s difficult to determine the combined incremental effect of these platforms without a unified approach.

Example Scenario – The Complexity of a Cross-Platform Campaign

Consider a company running a holiday marketing campaign across Facebook, Google Search, and Instagram. Each platform offers its own incrementality testing, but the methods and metrics differ. Facebook might show a 10% lift in sales, Google Search a 15% lift, and Instagram a 5% lift. Reconciling these numbers with the company’s overall sales increase is challenging, especially when internal data shows a different total increase. This discrepancy makes it difficult to optimize the budget allocation for future campaigns.

Alternative Options to Ad Platform Tools for Measuring Incrementality

Randomized Controlled Testing (RCT)

Randomized Controlled Testing (RCT) is often considered the gold standard for measuring incrementality. In an RCT, participants are randomly assigned to either the treatment group, which receives the marketing intervention, or the control group, which does not. This method allows for precise measurement of the campaign’s impact by comparing outcomes between the two groups.

Pros:

• High accuracy and reliability due to random assignment and control of external factors.

• Clear measurement of the causal effect of the marketing intervention.

Cons:

• Often impractical due to high costs and logistical complexity.

• Requires significant planning and coordination, making it challenging for smaller companies or those with limited budgets.

Marketing Mix Models (MMM)

Marketing Mix Models (MMM) are powerful tools used to understand the contribution of each paid channel to overall performance. They help allocate and optimize budgets and forecast future performance, aiding in budget justification and planning. MMMs, especially those using a Bayesian approach, require less data and use transaction data, avoiding the need for personal information.

Pros:

• Provides a comprehensive view of the effectiveness of all marketing channels.

• Helps in long-term strategic planning and budget allocation.

• Privacy-friendly as it uses aggregated transaction data rather than personal data.

Cons:

• More suited for longer-term, strategic analysis rather than immediate, tactical adjustments.

• Requires significant historical data and sophisticated modeling techniques.

GeoLift Experimentation

GeoLift experimentation is a popular and accessible method for measuring incrementality using geographic regions. This approach involves dividing markets into different geographic areas and running the marketing intervention in some areas while keeping others as control. Geo experiments require minimal data and are privacy-centered, making them less time and resource-intensive.

Pros:

• Minimal data requirements and privacy-friendly.

• Less time and resource-intensive compared to RCTs and MMMs.

• Provides clear and actionable insights for marketing strategies.

Cons:

• May not capture the full impact of nationwide or global campaigns.

• Requires careful selection and matching of geographic regions to ensure comparability.