Market Analytics Platform

10 January, 2024

The Future of Marketing: Why Media Mix Modelling is the Key to Success in 2024

Traditional methods of marketing attribution are becoming archaic, unable to keep pace with the complexities of modern consumer behaviour. This blog post delves into the the historical trajectory, current approaches, emerging trends, benefits, drawbacks, and practical insights for implementing attribution, specifically focusing on media modelling (MMM).

Why we care.

What’s the ROI of each of my marketing channels? What’s the optimal spend level for each channel to maximise KPIs? Does the way a campaign is executed on a channel impact its performance? Where should my ad spend go? These are just a few of the questions you’re probably asking yourself on a daily basis. Challengingly, the relationship between activity and results and the consumer journey is increasingly complex, so answering these fundamental questions is no easy task.

The Downfall of “Attribution”

Attribution models such as last-click were once effective, especially with digital channels.

However, expectations and market changes in 2023 revealed them as simplistic and increasingly unreliable. They fail to consider the multifaceted journey of today’s consumer, who interacts with brands through a plethora of channels. This narrow focus leads to skewed insights, making it nearly impossible for marketers to accurately assess the effectiveness of their campaigns. But what’s the alternative?

It’s time to turn our gaze towards a more robust, data-driven approach: Media Mix Modelling (MMM) combined with AI and machine learning.

The Rise of Media Mix Modelling

What is MMM?

MMM is a top down, statistical approach using historical data to predict factors affecting sales or revenue. It can explain the user acquisition saturation of a channel and serves as a future budget optimisation tool.

As a method, it analyses historical data to determine the impact of various marketing channels on sales and other performance metrics. In 2024, MMM has evolved and significantly improved both in terms of results and accessibility, primarily thanks to the integration of AI and machine learning algorithms. This has empowered MMM to handle massive datasets, uncover deep insights, and predict future market trends with unprecedented accuracy.

Background of MMM: From Inception to Evolution

Marketing Mix Modelling traces its roots back to the mid-20th century when marketers sought ways to quantify the effects of various marketing inputs on sales. Early models primarily focused on simplistic linear regressions, analysing the impact of factors like advertising spend, distribution, pricing, and promotions on sales figures.

Over time, advancements in technology, data analytics, and statistical methodologies paved the way for more sophisticated MMM models. The late 20th century witnessed the advent of econometric models followed by the integration of digital data and the rise of machine learning and AI-driven MMM models in the 21st century.

Both MMM and econometrics offer valuable insights into cause-and-effect relationships within the realm of marketing and business. While MMM provides a targeted analysis of marketing activities, econometrics offers a more holistic view that incorporates a broader range of economic variables and principles beyond marketing alone. Integration of these approaches can offer a comprehensive understanding of business dynamics and optimise decision-making processes.

MMM and Brand Strategy

The true potential of MMM lies in its ability to better evaluate the impact of efforts such as brand strategy, offline campaigns, or earned media. By providing a holistic view of marketing performance across all channels, MMM allows businesses to craft strategies that factor in the impact of the funnel, ensuring every marketing dollar is spent wisely.

Current Approaches and Trends

  • Data Integration and Granularity: Modern MMM approaches embrace diverse data sources, integrating traditional and digital marketing data, customer behaviour, market trends, and external factors like economic indicators. Granular data at individual customer levels or micro-segments offers deeper insights.
  • Advanced Analytics and Machine Learning: Sophisticated algorithms and machine learning techniques now power MMM, enabling predictive modelling, scenario planning, and real-time adjustments, fostering agility in decision-making.
  • Attribution and Multi-Touch Analysis: Shifts towards multi-touch attribution acknowledge the complex customer journey, moving beyond last-touch attribution models to capture the influence of multiple touchpoints on conversions.
  • Integration of Non-Traditional Channels: MMM has expanded beyond traditional advertising channels to encompass influencer marketing, social media, content marketing, and even experiential marketing, providing a holistic view of marketing investments.

Case Studies and Examples: Success Stories in the Making

Several companies have already reaped the benefits of this approach. For instance, a major retail brands like Klarna utilised MMM and Bayesian models to optimise online and offline advertising spend, resulting in a 20% increase in ROI. Whilst pet company BARK utilised MMM to fine-tune budget allocation, using MMM from Facebook Robyn.

Benefits and Drawbacks of MMM


  • Optimised Resource Allocation: MMM aids in allocating marketing budgets efficiently, maximising ROI by identifying high-impact channels.
  • Strategic Decision-Making: Provides actionable insights for strategic decisions, enabling informed marketing strategies aligned with business objectives.
  • Performance Evaluation: Facilitates the evaluation of marketing performance and effectiveness, fostering continuous improvement.


  • Data Complexity: Managing and integrating diverse datasets can be complex and resource-intensive, requiring specialised skills and technology.
  • Dynamic Market Dynamics: Rapidly changing market dynamics and external factors may render historical data less relevant, impacting model accuracy.

Emerging Shifts and Opportunities

  • Real-Time Decision-Making: Real-time MMM capabilities enable marketers to respond swiftly to market changes, adjusting strategies on-the-go for optimal results.
  • Predictive Modelling and Forecasting: Predictive MMM models forecast future scenarios, guiding proactive decision-making and strategic planning.
  • Cross-Channel Integration: The convergence of offline and online channels within MMM models provides a holistic view, aligning marketing efforts across touch-points for a seamless customer experience.

Implementation and Recommendations

To implement MMM effectively, businesses should start by collecting and integrating data from all marketing channels. Investing in AI and machine learning technologies is crucial to analyse this data efficiently. Additionally, companies should focus on building a team with diverse skillsets, including data scientists, marketing analysts, and brand strategists.

Practical Advice for Implementing MMM

  • Define Clear Objectives: Clearly define business objectives and KPIs to ensure alignment with MMM goals.
  • Data Quality and Integration: Invest in data quality and integration tools to ensure a robust dataset for accurate modelling.
  • Talent and Technology: Employ skilled data scientists and analysts proficient in MMM methodologies and leverage advanced analytics tools.
  • Continuous Iteration: Embrace a culture of continuous learning and iteration, refining MMM models based on real-time insights.

Conclusion: Navigating the Future of Measurement for Your Brand

It’s clear that traditional attribution models are no longer sufficient. Media Mix Modelling, augmented by AI and machine learning, is not just the future – it’s the present. By embracing this approach, businesses can make informed decisions, maximise their marketing ROI, and stay ahead in the competitive market.

The trajectory of MMM continues to pivot towards more sophisticated, agile, and encompassing models. Businesses leveraging the power of advanced analytics and integrated data sources stand poised to navigate the complexities of modern marketing landscapes effectively. Implementing MMM isn’t just about embracing technology but fostering a data-driven culture that drives strategic decision-making, ensuring sustainable growth and competitive advantage in today’s dynamic markets.

As marketers, it’s time to evolve or be left behind. The choice is ours. Delivered by BoldLens.

Read more

Access Data Like Never Before with Boldlens’ ‘Explore’ 

Access Data Like Never Before with Boldlens’ ‘Explore’ 

Jamie Viljoen Just ask ChatGPT… or there’s Explore…  ‘Imagine having all the data you need right at your fingertips, ready…
Media Mix Modelling:  Why Today's Marketers Can't Afford to Ignore It

Media Mix Modelling:  Why Today's Marketers Can't Afford to Ignore It

Elena Docheva Forget one-size-fits-all marketing. Enhanced by new advancements in machine learning and artificial intelligence, Media Mix Modelling (MMM) can…
Say Goodbye to Confusion and Hello to BoldLens Tooltips!

Say Goodbye to Confusion and Hello to BoldLens Tooltips!

Understanding marketing metrics just got easier – introducing tooltips in BoldLens.  BoldLens offers a wealth of marketing performance data, which…