Zeta Global’s leading marketing platform unifies customer engagement and measurement across all channels, giving marketers an advantage on speed, control, and accuracy. Read below, or contact us directly to learn more.
Marketers have increasingly benefitted from programmatic marketing, which delivers three advantages over legacy approaches: speed, control, and measurement accuracy.
This article describes best practices for omnichannel measurement and how to use these techniques to improve how marketers spend their dollars.
1. Media Mix Modeling
What Is It? Media Mix Modeling is one of the oldest forms of measurement and planning. By analyzing the relationship of aggregate spend in different media, marketers can understand how shifting the spend from one tactic to another drives revenue.
How Does It Work? Often media mix analysis is done at a geographic level. For example, if marketers spend more money on TV advertising on the west coast, and more on mobile advertising on the east coast, they can analyze the regional sales trends to see the impact of advertising. Marketers can also use Media Mix Modeling to analyze how shifting spend from one publisher to another improves their results.
What’s the Drawback? Media Mix Modeling analysis is conducted at an aggregate level, it cannot account for recency, frequency or sequences of media exposure at an individual level, all of which have a significant impact on effectiveness. For this reason, most marketers also rely on attribution measurement.
What Is It? The goal of attribution is the same as Media Mix Modeling: to quantify the influence of media exposure on revenue. However, attribution analyzes campaign data at the event level (views, clicks, etc.). This more accurate measurement further enhances the key benefits of programmatic marketing — feeding even better results into the always-on performance feedback loop within the marketing platform.
How Does It Work? There are three key approaches to “attributing” a conversion:
- Single event-based attribution, often called last-touch attribution (LTA), is the least accurate method. Assigning 100% of the credit back to the consumer’s last touchpoint does not distinguish between the search engine navigation and the advertising that influences people to enter the keywords into the empty search box.
- Multiple-event-based attribution (often called Multi-touch attribution (MTA) enables greater precision and control. This approach assigns partial credit to each consumer touch before a conversion. In our scenario above, both the search engine and the video ad would receive credit. Most modern marketers rely on self-learning algorithms that adapt the importance of exposure over time. Some DSPs offer automatic budget reallocation based on the best performing media to deliver better results by streamlining the feedback loop for their clients.
What are the Drawbacks? One of the challenges of MTA is that attribution modeling companies (e.g., C3, Visual IQ, Google/Adometry) vary in the results they provide (e.g., C3 found Adometry gave >360x more credit to Google). Also, consumer exposure is often tied to a single device and may undervalue omnichannel consumer experiences without the use of a probabilistic identity graph that can measure activity across different devices for the same person or household at scale.
3. Panel-based Measurement
What Is It? Panel-based measurement analyzes the activity for a sample of consumers and households, and then parlays the impact across the entire campaign. This ensures a more accurate collection of information across all of the same consumer’s devices relative to attribution. Panel-based measurement also can capture metrics associated with brand awareness, affinity and purchase intent.
How Does It Work? Panels provide consumers an opt-in method (such as surveys) to track deterministic consumer activity. This gives marketers a trade-off in scale for an increase in precision. Marketers project the panel results onto the larger population exposed to the marketing campaigns. Some advanced DSPs now incorporate panels and can use survey results to optimize the inflight budget allocations.
What Are the Drawbacks? Given the small sample size of the panel, often there may not be sufficient statistics to make sure the measured results will be repeatable. Also, consumers who opt into a panel may not be representative of a marketer’s target audience.
Zeta Leads the Charge on the Path Forward: Incrementality Measurement
The next horizon of measurement is not just assigning appropriate credit to prior exposures (MTA), but accounting for how the consumer would have acted without advertising — called “incrementality measurement.” This innovative analysis measures the true lift of the media spend and the incremental results it drives, allowing marketers to understand what to do next – in real-time.
Modern marketers benefit from the AI-driven optimization built into integrated programmatic marketing platforms by closing the gap between measurement and engagement. These platforms not only provide enhanced accuracy of measurement and control but provide the best ways to engage each brand’s target audience.