Quantified Nation #8: Driving business growth with marketing measurement
In today's competitive landscape, data-driven marketing is essential for driving significant business growth. Then, why is takes so much to have a proper marketing measurement practice in place? How teams play the game of making data-driven marketing decisions? What happens if some decisions are left to other types of intelligence?
The core of data-driven marketing lies in our ability to measure performance. This enables better decisions on channel budgeting and optimization, reduces wasted spend, and enhances campaign effectiveness. Ultimately, we aim to develop marketing and sales strategies that outperform our competitors in revenue, profit, or any financial metric you choose. Accurate measurement of the impact that each marketing investment has on financial metrics is the first step toward becoming a successful marketer.
The next step is to leverage these measurements to make decisions that drive higher growth. In marketing, our goal is to influence customer behaviors that boost our business—by increasing the number of buyers or enhancing their willingness to pay more. By accurately measuring these behavior changes, we can make informed decisions that significantly boost our business success.
Measuring how digital marketing changes behaviors
Digital marketing allows us to target any online user, tailoring the creative, message, and timing for maximum impact. By measuring various aspects of our marketing efforts—such as creativity, messaging, and delivery methods—with high granularity and in real-time, we can effectively analyze their impact on user behavior and business outcomes. This precision and timeliness in measurement enable us to make better, faster decisions and continually refine our strategies for maximum growth.
Making good decisions with vast amounts of data has always been a challenge. Today, many of these decisions are made by algorithms and models, often referred to as AI. These AI systems process data to make precise and efficient marketing decisions.
The Role of Artificial and Human Intelligence
Today, much of digital marketing is automated, with AI systems optimizing campaigns in real-time. These advanced systems handle data seamlessly, deciding which ads to show to specific users and how much to pay for each impression, click, or conversion. While AI manages these tasks, humans play a crucial role in setting the overall strategy. They allocate budgets and set goals based on continuous data flows from user interactions with the company's marketing efforts and sales channels.
By combining human strategic decisions with AI's data processing and real-time optimization, companies can achieve highly profitable marketing budgets.
Navigating the Complexities of Digital Marketing Measurement
Data feeding AI systems is primarily based on digital tracking, widely adopted with best practices. However, superior performance requires effort and dedication due to ecosystem changes and system complexities. Transforming this information into better AI-driven decisions involves extensive testing and predictions. Mastering this process is a major competitive advantage for companies, even bigger for those with large marketing budgets.
Companies aim to optimize better and faster than competitors, making better data-driven decisions at strategic and operational levels. Ensuring marketing measurement feeds AI systems with the best available data is crucial.
Current Status of Digital Tracking
Recent ecosystem changes have reduced the ability to collect relevant marketing data. These changes include third-party cookie elimination in Safari and Firefox (with Chrome soon to follow), consent requirements (which we discussed in Quantified Nation #6), and private browsing modes or tools. These are just a few examples, with many more changes to come.
Digital marketers must adapt, understanding the complexities faced by their analytics teams. Embrace a new context where precision is diminished, and modeling becomes the norm, though not the holy grail. We must master training AI systems in this new ecosystem which requires additional ways to send data to the platforms and deeply review how teams make decisions using this data.
Strategies for Effective Measurement
Balancing the need for high-quality data in marketing AI systems with its reduced availability is crucial. Here's how you can measure better than your competitors:
Understand Data Gaps: Identify gaps in your measurement system and consider them when making decisions or setting up automations. Adjust goals and targets based on these gaps and monitor accordingly.
Monitor closely your data collection: When you have AI systems and people making decisions over your business results, you need tools to ensure the good flow of data through the whole process.
Model Data Gaps: Use other marketing measurement methodologies, such as incrementality or MMM, to model data gaps. This process is ongoing, so strive for continuous improvement, not perfection.
Embrace Comprehensive Metrics: Include metrics beyond conversions, such as brand reach and frequency. Link these metrics to financial benchmarks for context-based decision-making.
Empower Your Teams: Enable your teams to make decisions based on diverse information sources. Challenge them to develop and test new hypotheses continuously.
While no marketing measurement system is perfect, strive to ensure yours is better than your competitors'.
Drive business growth by feed your AI systems with the best data for optimization and ensure your teams make the best decisions with a robust measurement strategy.
Bayesians vs Frequentists
Did you know the Marketing Mix Model library from Meta (Robyn) uses frequentists statistical methods while the one from Google (Meridian) uses Bayesian statistical methods? It turns out there is a strong debate across providers, some defending their frequentist approach such as Aryma Labs and others their bayesian approach such as Recast. Where to start? The amazing mathematics YouTube channel 3Blue1Brown has a great explanatory video on Bayes’ theorem.
Industry updates and upcoming events
The WFA (World Federation of Advertisers) recently hosted a summit on Halo, their overall framework for cross-media-measurement (details here). This happens as the Halo-inspired projects in the UK (Project Origin) and the USA are approaching beta testing. You can find a summary of the key themes of the summit on this link.
Kantar Media has just released a document on the role of media consumption panels (access on this link). They talk about the importance of people-based measurement from several perspectives (privacy, accuracy…). The most interesting piece, in our opinion, was the part on how to link 1P data from media providers with the data from Kantar’s panel, allowing for targeting, calibration, sales activation… Specific examples with different media companies and trade bodies are shared. We spoke about these topics also in Quantified Nation #2.
The ARF (Advertising Research Foundation) has published an update on their “Attention Measurement Validation Initiative” (link). This latest report focuses on best practices on creative testing using attention metrics. In order to download the report, membership to the association is required.
Chart of the week
Is summer coming earlier? Complex question. What we can say for sure is that people’s queries on “summer” are happening earlier on this 2024. In this chart you can see that the pattern of 2010 (yellow) and 2018 (blue) is very similar, but the 2024 queries (red) started to raise several weeks earlier.
Oldies but goodies
Soon it will be 4 years since the release of Google’s Decoding Decisions research. It explored the complexity of decision-making in the internet age through the lens of behavioral sciences. A full length report is available for download on this link.
