Marketing Mix Modeling, News • 7 min reading time

Making media measurement actionable

Brian Cusick - published on November 2, 2023

People seek out media measurement for many reasons. Of course, the primary reason is self-evident in the name, but the specific motivations, the priorities, and the expectations can vary wildly. Media measurement can be defined as the practice of quantifying the impact of media on your KPI. But measurement can cover many dimensions, so it should align with an organization’s goals, the extent of media coverage and an organization’s willingness to formulate and execute plans based on the output. We often see RFPs that read like a laundry list of everything one could ever want from a measurement system. This is typically a red flag: “give me everything, make it fast, cheap, and perfectly reliable” is not a serious conversation. It turns out that to make media measurement actionable one must create real priorities. Priorities identify what is more important, and this is what typically comes to mind when one thinks about prioritization, but they also identify what can be sacrificed if need be, and in business, an honest assessment of what is less important is often more valuable than identifying what is most important (if for no other reason than the latter is typically more obvious).

Will measurement give you everything you want? 

Probably not. Good measurement will only give you what the data set supports; statistical methods are beholden to the theory behind those statistical methods. Methods/systems that appear to answer all questions are too often smoke and mirrors – they either fill in the gaps with hard coded prior beliefs, or they employ some form of “fair share”- neither of which are actual measurements. The best way to get as close to what you want is to draft out your measurement system in reverse:    

  • Start with the planning view, then let the experts tell you the models required to support that planning 
  • Set data requirements based on the required models 
  • If you can’t get reliable data, go back to the beginning and change expectations for planning, which, in turn, will change the models, and in turn, the data requirements  
  • Remember, over time most organizations can improve the data that is generated and/or made available and the larger process can then be revised and improved and can get closer and closer to meeting your ultimate goals (i.e., your wish list) 

One might be surprised by how common it is for this to work in the reverse order. The first question we get is sometimes, “What data do we need?” with the expectation that once we have the data, we will build the models and then inform planning. Designing measurement processes in this order risks final output that is hard for the team to use and limits the actionable outcomes. 

Consider what this means in practice  

Is the goal to inform budget allocation?  

  • Between brands or between markets? 
  • Between spends designed to drive different levels of the marketing funnel? 
  • Between tactics, campaigns or individual creative spots?  

Is the goal to understand drivers of revenue growth?  

  • Growth drivers will be broader than just paid media and will differ based on the industry, including, price, distribution, macro-economic factors, etc. 
  • Will the Marketing Mix Model be driving a forecast of total sales?  A true forecast is very different than a budget allocation recommendation. 

At what cadence do you build and change these plans? 

  • Annual or multi-year strategic planning? 
  • Monthly forecast updates? 
  • Daily or weekly execution / budget optimization? 

At ScanmarQED, we help our clients to work through these questions. We offer: 

  • Multiple media measurement platforms that support different sorts of planning needs  
  • A modern approach to data management  
  • A team of true experts to design and support the entire process 
  • Our tech is scalable, reliable, and affordable but our recommendations are based on meeting your needs, not pushing our agenda or insisting on “only 1 correct way”    
  • We’ve guided both large and small organizations through this journey, from multi-brand global giants with $100M+ in working media to single brand-market entities with as little as $1M in annual media spend.  

Get personalized assistance

Contact us to discuss how we can help your organization.

Picture of Brian Cusick

Brian Cusick

Brian Cusick leads the North American business unit for ScanmarQED, ensuring that both new and existing clients get the most value from our services and products. His expertise is in quantitative marketing research (previously worked at AC Nielsen, PepsiCo, Kraft Foods), with a focus on data driven business planning and marketing mix based optimization.