Marketing Attribution Models: Shapley, Markov, and Beyond

In the dynamic realm of digital marketing, the quest for understanding the impact of marketing channels on conversions has led to the development and use of various attribution models. Among these models, Shapley and Markov are two of the most commonly used marketing attribution models. In this post, we will embark on the strengths and weaknesses of Shapley and Markov models and look for a new solution that is more adaptable to modern-day marketing.??

Shapley, named after the Nobel laureate Lloyd Shapley, and Markov, inspired by the mathematician Andrey Markov, have roots in different scientific disciplines but found their application in marketing attribution due to their mathematical underpinnings. Shapley's work on cooperative game theory and Markov's contributions to probability theory were pivotal in shaping these attribution models.

How Shapley Attribution Works

Shapley's attribution revolves around the concept of fairness. It assigns credit to marketing touchpoints based on their position within the customer journey. Each touchpoint's contribution is calculated by considering all possible permutations in which it appears, rewarding touchpoints that are more likely to lead to conversions.

Strengths of Shapley Attribution:

• Simplicity and ease of understanding.
• Fairly distributed credit among touchpoints
• Suitable for relatively linear customer journeys.

Weak Points of Shapley Attribution:

• Struggles with complex, multi-channel interactions.
• Ignores the timing and sequence of touchpoints.
• It may not reflect real-world customer behaviour accurately.

Decoding Markov Attribution

Markov models, on the other hand, provide a more holistic view of the customer journey. They employ probability theory to calculate the likelihood of a conversion occurring based on the path of touchpoints leading up to it. This approach considers the sequential nature of interactions.

Strengths of Markov Attribution:

• Considers the entire customer journey.
• Accounts for timing and sequence of touchpoints.
• Provides a probabilistic view of attribution.

Weak Points of Markov Attribution:

• Requires large amounts of historical data.
• Complexity in modelling.
• Limited ability to handle highly convoluted customer journeys.

The Dawn of Marketing Spend Optimisation (MSO)

In contrast to traditional models, MSO represents a leap forward in precision and utility. It begins with recognising that every customer's journey is unique and that channels do not contribute equally to conversions. MSO combines customer journey data with marketing spending information, allowing for precise attribution and optimisation.

The Advantages of MSO:

• Precise Attribution: MSO accurately attributes conversions to specific channels.
• Budget Allocation Without Increased Spending: It empowers marketers to optimise ROI without expanding budgets.
• Simulation for Maximum Conversions: MSO employs simulations to identify optimal budget distribution.
• Addressing Pain Points: MSO provides solutions for wasted budget, low conversions, and missed opportunities.
• Adaptation and Refinement: Insights from MSO help marketers adapt and refine their strategies in real time.

In conclusion, while Shapley and Markov attribution models have played significant roles in understanding marketing effectiveness, the digital landscape has evolved. The intricacies of modern customer journeys demand a more precise, data-backed approach, and MSO stands at the forefront of this evolution. It's not about dismissing the historical models but rather embracing a more comprehensive and accurate solution that ensures marketing efforts are recognised and optimised to their full potential in the ever-changing world of digital marketing.