[Podcast] Episode 10: AI, Surveys, and Regional Tactics for CMOs

Are you overwhelmed by the sheer volume of data from various marketing platforms and methodologies? It’s now harder than ever to discern ‘true effectiveness‘.

In our latest episode, we dive into the intricacies of modern marketing, offering practical strategies to cut through the data clutter and achieve marketing clarity. Our hosts will cover how you can leverage AI in Marketing Mix Modelling to streamline data processing. Learn the importance of maintaining high-quality, consistent data for accurate insights, and explore the benefits of using surveys to capture insights from smaller channels, as well as gaining insights into regional hot housing and testing to measure the impact of specific campaigns.

Watch the full episode or read the blog below.

AI, Surveys, and Regional Tactics for CMOs

Brands have access to more data points than ever before, yet the challenge of measuring true effectiveness persists. With various attribution methodologies and platforms offering their own versions of success metrics, understanding what works and why – can be difficult to pinpoint.

The Evolving Role of Data in Marketing

Data has become a central force in marketing, with many viewing it as the key to unlocking better campaign performance and return on investment (ROI). However, there is no single source of truth that marketers can rely on to gauge effectiveness. Instead, success lies in adopting a multi-faceted approach—using various tools and methodologies tailored to specific business needs. By understanding the broad landscape, marketers can choose the right tools for different situations and make sure they are not overly reliant on any one solution.

Defining Clear Objectives: The Foundation of Effective Measurement

A significant part of the measurement challenge comes from not clearly defining objectives. When a brand asks how to achieve a specific ROI, the real question often lies beneath the surface:

What are the core objectives?

Only by answering such a question can the right measurement technique be selected, whether they involve sophisticated attribution models or basic numerical analysis.

Despite this, you need to be comfortable with uncertainty or the unknown. In data science, much of what is measured revolves around probabilities and confidence intervals rather than fixed truths. To find actionable insights, marketers must embrace the unknown and learn to make decisions based on a range of possible outcomes.

The Shift from Single-Source to Multi-Tool Measurement

Traditionally, platforms like Google Analytics were considered a gold standard for measuring marketing success. However, as tools have evolved, it’s become clear that no single platform can offer a complete view. Google’s shift to data-driven attribution modelling in Google Analytics 4, for example, has shown some limitations in accuracy, particularly when compared to the older Universal Analytics. While these platforms remain important, it’s critical for brands to diversify their measurement approach, incorporating multiple tools to create a more comprehensive understanding of campaign effectiveness.

Macro vs. Micro Measurement Approaches

One effective way to navigate the data measurement landscape is by combining macro and micro approaches. Broadly speaking, macro approaches provide a big-picture view of performance, which can help with strategic decisions while micro methods offer granular insights into individual campaigns and will help in understanding the why.

Marketing Mix Modelling (MMM)

At the macro level, Marketing Mix Modelling (MMM) offers a way to quantify the impact of marketing across different channels. By using data across several years, MMM can isolate the effects of marketing activities on key performance indicators (KPIs) like sales or revenue, while controlling for external factors such as seasonality or economic shifts.

MMM is particularly useful for brands with omnichannel campaigns, as it allows for a holistic view across both online and offline channels. However, one limitation of MMM is that it doesn’t offer the granularity needed to explain why a campaign worked—only that it did. Additionally, MMM is time-intensive, often requiring quarterly updates, though advancements in technology are allowing for faster updates, making monthly reports more feasible in some cases.

Micro Channel-Specific Measurement

On the micro side, channel-specific tools can provide deeper insights into individual campaign performance. For instance, survey attribution—such as “How Did You Hear About Us?”— can capture feedback directly from customers. This method helps bridge the gap between deterministic digital models and probabilistic offline attribution by applying a single methodology across all media channels.

While survey data provides valuable insight, it’s essential to handle it with care to avoid bias. For example, rotating answer options or including dummy responses can help ensure more reliable data. Alongside that, combining survey responses with last-click data can offer a more unified view of a campaign’s impact across multiple channels.

Overcoming Data Quality Challenges

One of the biggest challenges in measurement is data quality. Whether using MMM, micro attribution, or surveys, the accuracy of insights depends on the consistency and reliability of the underlying data. In many cases, this means collecting at least two to three years’ worth of data, and making sure there’s consistency in methodologies, as well as having robust data infrastructure in place.

Changes in data sources, like shifting suppliers or platforms, can disrupt the modelling process. Additionally, not all advertisers have access to large datasets, which makes it harder to measure campaigns accurately, especially when seasonal factors need to be accounted for. Careful attention to these data quality issues is crucial for producing actionable insights.

AI and the Future of Measurement

Artificial intelligence (AI) is poised to change how data is processed and insights are generated. Platforms like Google and Meta are already developing automated modelling tools that aim to simplify the measurement process. While AI can reduce the time and effort involved in modelling, the strategic human element remains essential. AI can quickly generate reports, but it takes an expert understanding to interpret those numbers and turn them into meaningful, actionable insights.

Regional Testing: A Targeted Approach to Attribution

One approach gaining traction is regional testing or regional hot-housing. This method helps brands assess the effectiveness of harder-to-measure channels, such as radio and out-of-home (OOH) advertising. By running campaigns in specific regions and comparing results to a control region where no activity is taking place, marketers can get a clearer view of channel performance.

Regional testing works best when control and test areas are similar in size and audience makeup. For example, London is often excluded as a test region because it’s difficult to find a comparable control area. Instead, focusing on smaller, more similar regions can provide more accurate results.

Best Practices for Measurement

To make the most of data and measurement, brands should consider the following best practices:

• Use multiple tools: Don’t rely on a single platform or methodology. Combine macro models like MMM with micro-level insights from tools like survey attribution.
• Ensure data quality: Consistent, long-term data is essential for accurate modelling. Invest in data infrastructure and processes to maintain high-quality data.
• Consider regional testing: For omnichannel campaigns, regional testing can provide valuable insights into hard-to-measure channels.
• Leverage AI smartly: Use AI to streamline processes but retain human expertise to drive strategic decision-making.
• Tailor methods to your brand: Smaller brands with fewer channels may not benefit from all measurement techniques. Focus on the tools that align with your business needs.

With so many tools and methodologies available currently, brands must take a holistic approach, using both macro and micro tools to gain a full understanding of their campaigns’ performance. Data Science offers powerful insights, but only when the right questions are asked, the right tools are used, and the data itself is of high quality. By understanding the strengths and limitations of various measurement tools —whether macro models like MMM or micro approaches like survey attribution —advertisers can make more informed decisions about where to invest their marketing spend.

Get in touch with our digital team today

Find out how we can combine digital, TV and data science expertise to get the most out of your advertising budget.