Is there a silver bullet for marketing attribution?


Share on LinkedIn

The marketing attribution landscape has evolved over the past decade. While historically centered around marketing mix modeling, it has recently become focused more on bottom-up digital attribution as a stand-alone solution. It has become clear that marketers are looking for continued evolution toward a more balanced approach utilizing multiple attribution methods. In short, they have realized there is no silver bullet to attribution.

A recent Forrester report on cross-channel attribution has highlighted this dissatisfaction on the part of marketers. Attribution vendors’ net promoter scores went from 71 to 29 over a two-year period (from 2012 to 2014). Marketers are changing attribution vendors more frequently, with the majority having their current vendor for less than a year. Marketers are unhappy with optimization recommendations, are looking for measurement beyond just digital, and want to see more transparency in the solutions.

While the optimal measurement roadmap is different for each marketer, there are some common tenets that define the shift in thinking.

  • Accurate measurement requires a holistic framework, including offline as well as online media. It has been proven that offline media drives consumption of online media, especially search. Therefore, any attribution solution measuring digital without taking into account the offline impacts will result in biased estimates and misallocation of resources.
  • While marketing mix modeling has been around for decades and has been proven across industries, digital attribution is relatively new and should be “pressure tested” before being used. Marketers are now viewing digital attribution results with skepticism, when it often shows apparent bias – for example, when retargeting is the highest performing digital tactic, with no consideration for the funnel dynamics.
  • Focus on the data – garbage in, garbage out. Anyone with experience in marketing mix modeling knows that a majority of time is spent on understanding the data trends for validation and informing more relevant analytics and insights. The size and standardized layout of digital data has enabled digital attribution solutions to become very automated in terms of data ingestion, which has led to less focus on data quality validation. As a result, the risk of attribution based on inaccurate data has grown.
  • Reporting and optimization tools can add value, but should not replace strategic analytic support. These tools are great at automating repeatable processes and insights, however marketers often require custom analytics to address evolving business needs and drive additional value.
Brian Demitros
Brian has 11 years of experience in marketing analytics and econometric modeling, primarily focused on leading the design and implementation of Attribution & Media Mix projects for Fortune 500 companies including Walt Disney, GEICO, Wal-Mart, AT&T, Walgreens, Chrysler, ConAgra, Avon and Visa. Prior to joining Merkle, Brian worked for the SymphonyIRI Group performing pricing modeling and Media Mix analysis and Accenture Interactive performing Media Mix modeling. Brian has a MS degree in Applied Economics from Marquette University.


  1. We’ve seen this prove true in our own conversations with clients. I wonder how much of the “debate” between MMM and bottom-up/tag-based attribution (which should coexist happily) is the result of marketing execs who are coming up through marketing ranks rather than thinking about the broader picture that includes finance, sales and other functional considerations.

    Entire organizations benefit from MMM/econometric modeling, whereas only marketing managers and creative teams benefit from granular attribution, and then typically only the digital side.

    I’ve been impressed by the way C3 Metrics approaches “bottom-up” attribution, with omni-channel capabilities and a very intelligent application of customer journey philosophy.

    It’d be great to talk more about MMM Brian, as well as agile analytics/”IT for marketing” services.

    My email: [email protected]


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here