In the digital age, the marketing landscape is rapidly evolving. Not long ago, marketers thrived on extensive data collection methods that delved deeply into consumer behaviours and preferences, often pushing the boundaries of privacy. Today, the environment is drastically different. Heightened consumer privacy expectations coupled with stringent regulations like GDPR and DMA are compelling marketers to rethink their strategies. These changes are driving the industry towards more ethical practices that prioritise consumer privacy without compromising the ability to measure and optimise marketing effectiveness. Inspired by comprehensive resources such as Google’s The Modern Measurement Playbook (which, imho, is a must read for anyone that is working in digital advertising), this article outlines a strategic framework for adopting new measurement methodologies that not only comply with these evolving standards but also enhance strategic advantage. We will explore when and how various types of businesses can implement these modern approaches to thrive in this new era of marketing.
The Need for Change in Marketing Measurement
As digital privacy becomes a global focus companies are required to adapt their marketing strategies. This adaptation involves more than just compliance; it’s about shifting towards a focus on first-party data due to restrictions on broader data collection practices.
First-party data is increasingly vital not only for adhering to legal standards but also for maintaining effective marketing strategies in a privacy-focused world. Utilising new measurement techniques such as Consent mode V2 and easier than ever server-side tagging is crucial for navigating these challenges.
These methods help ensure that marketing efforts are both ethical and effective by relying on data directly gathered from customer interactions, which is transparent and privacy-compliant. This approach strengthens consumer trust and supports the sustained effectiveness of marketing campaigns.
Implementing “New” Methodologies
The path forward involves integrating advanced, privacy-friendly measurement techniques that provide valuable insights while respecting user consent. These methodologies not only align with regulatory requirements but also offer strategic advantages by improving the quality of insights derived from marketing data.
Custom Attribution Model
Who Needs It:
Businesses engaged in e-commerce or any form of online selling stand to benefit from having their own custom attribution model. These models are crucial for companies that can track direct responses to their marketing efforts, such as clicks leading to purchases.
Strategic Importance:
Custom attribution provides a granular view of which marketing activities contribute directly to sales. This method allows businesses to allocate their marketing budget more effectively, ensuring that each dollar spent maximises return on investment. It forms the basis for understanding direct returns and planning budgets with greater precision.
Explanation:
Custom attribution models are designed to assign credit to different touchpoints in a customer’s journey that lead to a conversion. Unlike simpler models, custom (data-driven) attribution allows businesses to create a more nuanced understanding of how various marketing efforts contribute to final outcomes.
Requirements:
User-Level Data: The foundation of custom attribution is detailed user-level data that tracks entire customer journeys across multiple touchpoints and channels.
Properly Defined Source/Mediums (or any other dimension you’ll use when creating a model): Clear labelling and consistent definitions of sources and mediums are essential for accurately attributing conversions back to the right campaigns and channels.
Modelling Capability: Businesses can use various models depending on their complexity and data maturity. While data-driven models based on algorithms like the Markov chain are preferred for their ability to fairly distribute credit across touchpoints, simpler rule-based models like last-click or linear are also viable.
For instance, an online retailer could implement a custom attribution model to track the effectiveness of their multi-channel campaign, which might include email marketing, PPC ads, and social media. By using a data-driven attribution model, the retailer can see that while PPC ads are driving immediate purchases, their email marketing campaigns are effective in gradually converting interest into sales over time.
Marketing Mix Modeling (MMM)
Who Needs It:
MMM is vital for companies that operate complex advertising ecosystems involving multiple channels and diverse strategies. It’s particularly crucial for organisations focused on extensive branding efforts or those engaging in significant upper-funnel advertising activities (basically channels where direct sales isn’t expected, think branding ads on Google Display Network, YouTube ads, Facebook awareness ads, etc.). Additionally, MMM is beneficial for businesses that need to balance the direct impacts of online sales with broader goals of enhancing brand awareness. As marketing strategies evolve and expand, understanding the holistic performance of these multifaceted advertising mixes becomes critical, making MMM an indispensable tool for discerning the optimal configuration of marketing efforts.
Strategic Importance:
MMM is crucial for providing a comprehensive overview of the effectiveness of each marketing channel and tactic in a company’s advertising mix. As businesses expand their use of diverse marketing channels, the potential for cannibalization increases — where one channel’s gain in effectiveness comes at the expense of another. MMM addresses this by offering insights that help distinguish between the incremental impact of individual channels and the overlap between them.
This modelling is especially valuable in scenarios where multiple channels contribute to the same business outcomes, as it helps in attributing value accurately and preventing the misallocation of marketing resources. By delivering a holistic view of all marketing activities, MMM enables businesses to understand not just the direct effects of individual campaigns but also their combined influence on overall sales and brand equity. This holistic perspective is vital for making informed decisions that optimise both short-term results and long-term brand development, ensuring that marketing budgets are invested in ways that genuinely enhance brand value and market position without unnecessary redundancy and inefficiency in spending.
Requirements:
Aggregated Data: Unlike custom attribution that relies on user-level data, MMM utilises aggregated data. This data doesn’t track individual user behaviours but rather analyses overall trends and outcomes, making it well-suited for high-level strategic insights. TL:DR no user level data, means that it’s GDPR friendly.
Infrastructure and Expertise: Effective implementation of MMM requires a robust infrastructure (which is now rather easy and straightforward to build with cloud based solutions, like GCP) and access to data engineers (ideally ones that have marketing background, or at least understand how different marketing channels work). These professionals run complex models, like Facebook’s Robyn or Google’s Meridian, and are crucial for interpreting the results accurately. This technical capability enables businesses to refine their marketing strategies based on reliable, data-driven insights.
Lift Tests
Who Needs It:
Ideally, all businesses should engage in lift testing to validate the effectiveness of specific marketing campaigns or tactics. However, the frequency and scale of these tests can vary significantly based on the size of the business and the maturity of its marketing efforts.
Strategic Importance:
Lift tests allow businesses to experiment with and measure the incremental impact of marketing changes in a controlled setting. For smaller companies or those new to market, running lift tests annually can provide critical insights without overwhelming limited resources. Larger companies or those in more competitive markets might conduct these tests more frequently to stay agile and responsive to market dynamics.
Requirements:
Basic Setup: Simple lift tests can be conducted directly within platforms like Facebook and Google Ads, making them accessible even to smaller advertisers.
Advanced Planning and Execution: More sophisticated lift tests, suitable for large clients, require careful planning and execution across all channels, potentially including strategies like “go dark” periods or targeted audience segmentation.
Integrating Custom Attribution, MMM, and Lift Tests into a Unified Strategy
Source: The modern measurement playbook: How to optimise your marketing effectiveness and fuel growth
Starting with Custom Attribution
Custom attribution should be the foundation of your daily marketing operations. It provides immediate, granular insights into how individual marketing touchpoints contribute to conversions. This data feeds directly into tactical decisions, such as adjusting ad spend on specific channels, tweaking campaign messages, or reallocating budget between performing and underperforming assets.
Process Implementation:
Data Collection and Analysis: Begin by ensuring that your data collection is robust, capturing all necessary customer journey touchpoints across channels. This involves setting up proper tracking to accurately attribute conversions to the correct sources and mediums.
Real-Time Application: Use insights from custom attribution to make real-time adjustments in campaign management. This dynamic approach allows for quick responses to performance metrics, optimising ROI from direct response campaigns.
Informed by Marketing Mix Modeling
While custom attribution handles the immediate allocation of budgets and adjustments, MMM should be used to understand the broader impacts of your marketing spend across various channels, including offline media. MMM helps in strategic decision-making by showing the long-term effectiveness of different marketing elements and their role in driving overall business objectives like brand awareness and market share growth.
Process Implementation:
Quarterly or Bi-Annual Reviews: Incorporate findings from MMM into higher-level strategic planning sessions. This might involve adjusting the marketing mix based on which channels show the highest total return or shifting strategies to focus more on brand-building activities if MMM indicates their long-term value.
Feedback Loop: Use insights from MMM to set parameters for custom attribution models and refine what and how you measure in shorter cycles, ensuring that both approaches align and inform each other.
Refinement through Lift Tests
Lift tests are critical for validating the assumptions and insights gained from both custom attribution and MMM. They provide empirical evidence about the effectiveness of specific marketing tactics or strategies, offering a way to test hypotheses generated by longer-term data analyses.
Process Implementation:
Scheduled Testing: Depending on the size and maturity of the business, schedule lift tests at strategic intervals—perhaps quarterly for dynamic markets or annually for more stable industries.
Cross-Channel Coordination: Coordinate lift tests across all channels to understand the incremental impact of marketing changes comprehensively. Also try and plan things around classical vertical seasonality, or known non-paid demand (e.g. take into account back to school, or Black Friday)
Creating a Learning Environment
Establish a continuous feedback loop where insights from lift tests feed back into both the custom attribution models and MMM efforts. This learning environment helps refine hypotheses, improve data accuracy, and enhance the overall marketing strategy.
Iterative Learning: Regularly update your models and strategies based on the latest test results and market conditions. This iterative process helps keep your marketing efforts both relevant and compliant with changing consumer behaviours and regulatory environments.
Documentation and Sharing: Maintain thorough documentation of outcomes and insights from all methodologies. Share these learnings across the organisation to align all stakeholders and improve cross-functional strategies.
Conclusion
The evolution towards more privacy-focused marketing measurement is not just a response to regulatory pressures; it’s a strategic shift that can lead to more sustainable and consumer-friendly marketing practices. By aligning custom attribution, marketing mix modelling, and lift tests within a unified framework, businesses can create a robust, data-driven marketing strategy that adapts to both immediate market conditions and long-term strategic goals.
This integrated approach not only maximises the effectiveness of marketing spend but also ensures that decisions are informed by a comprehensive understanding of consumer behaviour and market dynamics, a theme extensively explored in Google’s The Modern Measurement Playbook.
As the industry continues to evolve, those who adapt swiftly and smartly will likely find themselves at a competitive advantage, with a better understanding of their customers and a more robust approach to marketing in the digital age.
