Marketing Mix Modeling and Multi-Touch Attribution: Why a Multi-Model Approach is a Marketer’s Measurement Dream

By Wayne St. Amand On Mar 7, 2019

Marketing departments are no longer measured by the leads they generate, but by providing proof of ROI for every campaign they run. This makes the need for accurate measurement increasingly important.

Yet traditional marketing measurement approaches such as last touch are unreliable and imprecise, and often lead marketers to make the wrong decisions when it comes to budget allocation.

There are, however, two distinctly different measurement methods that offer marketers advanced ways to identify the channels, campaigns, and tactics that drive the best performance: marketing mix modeling and multi-touch attribution. Both of these approaches enable marketers to make smarter investment decisions, yet each has its own unique advantages.

Plan Strategically with Marketing Mix Modeling

Marketing mix modeling provides high-level strategic insights for budgeting and long-term planning. It uses statistical analysis to understand past trends and predict the future effect of marketing tactics on sales. Leveraging summary-level data, it measures a wide array of online and offline marketing activities (linear TV, in-store promotions, online display, paid search, etc.), and assesses the effectiveness of spending by channel over and above the baseline of sales that would have occurred without any marketing efforts. It also controls for external factors that can impact performance, such as weather, economic conditions, and holidays.

Optimize Tactically with Multi-Touch Attribution

Multi-touch attribution, on the other hand, offers tactical insights for short-term optimization. It focuses on addressable channels and leverages granular, user-level data to analyze performance in near real time. It does this by calculating and assigning fractional KPI credit to the marketing touchpoints and dimensions (publisher, placement, creative, offer, etc.) along the consumer journey that influenced a sales or other desired outcome. Marketers can use this insight to make smarter tactical decisions, such as which call to action to use or which keywords to bid on.

When to Use Each Approach

While both measurement methods enable marketers to make more informed decisions, they each have their own strengths and work on their own timeline. Marketers have different needs for information based on their role within the department and the specific challenges they are facing, so understanding when to use each approach is critical.

The first thing a marketer needs to consider is how often they need the data to refresh. Multi-touch attribution is best used for pinpointing the channels and tactics that are most (or least) effective and reallocating budget while campaigns are still in-market. It provides insights at a much faster cadence than marketing mix modeling because models can be rebuilt daily and applied to each new day’s performance data. By revealing the nuances of what drove each individual day’s conversions, marketers can make informed decisions based on the most accurate and up-to-date information.

Another factor to consider is coverage. Marketing mix modeling is best for delivering broad recommendations for how marketers should allocate their budgets across both online and offline channels. Unlike multi-touch attribution, which is limited to addressable channels, marketing mix models can incorporate a marketer’s entire market mix.

Achieve Better Results by Combining Methods

Marketers’ need for an integrated view of their marketing performance is as critical as ever. While marketing mix modeling and multi-touch attribution have their distinct advantages, marketers can realize the highest returns when they are used in tandem. By bringing insights from both methods of analysis together, marketers can get the comprehensive view of performance they need to inform their next best action within and across channels.

For example, consider a brand that is running multiple campaigns across a number of online and offline channels. Using marketing mix modeling, the brand can analyze all the drivers of performance and determine how to best allocate its budget between channels on a quarterly or annual basis. Using multi-touch attribution, the brand can then drill down deeper and allocate channel budgets to the highest-performing publishers, placements, keywords, creatives, and other tactics while campaigns are still in flight. Armed with this strategic -and tactical-level insight, brands can make a broad array of decisions to maximize the efficiency and effectiveness of their entire marketing portfolios.

Most brands today rely on both online and offline channels to drives sales, revenue and other desired business outcomes. By leveraging a holistic measurement approach that reconciles data and insight from marketing mix modeling and multi-touch attribution, marketers can enjoy the truly comprehensive view they need to make the right decisions, at the right time, to optimize performance across every area of investment.