INside Performance Marketing
PMI Starts Today – Join Us There's still time to make it down to the Westminster Park Plaza Buy your ticket now
Making the Vital Shift to True Connected Attribution

Making the Vital Shift to True Connected Attribution

Measuring the incremental value marketing touchpoints have on consumer engagement and sales is a significant challenge facing marketing organisations today. A common issue is having multiple measurement solutions, none of which represents a complete picture of the marketing activity, and all of which produce different results and recommendations. It is common to see marketers trying to rationalise three different measurement systems:

  • A top-down measurement approach to understand performance across all media, competitive, and other external factors at a macro level (often marketing mix modelling, informing budget planning processes).
  • A bottom-up measurement approach to understand very specific digital media performance at a granular level (often a digital attribution solution operating on cookie-level data, informing tactical decisions within digital).
  • A simplified last touch or direct attribution solution feeding the primary accounting systems that set marketing goals 

None of these measurement approaches are sufficient – there is a need to solve this challenge in the market through a cohesive, holistic approach.

Bridging the gap

Over the past few years, significant effort has been made to bridge the gap between the top-down and bottom-up approaches. Digital attribution providers are making strides to integrate their systems with top-down modelling results or begin building top-down models themselves. Marketing mix providers are either partnering with digital attribution providers or beginning to build out their own bottom-up capabilities. However, the connection back to advertisers’ internal reporting systems is still lacking, and without that final integration, the vision of the single view of the truth will never be realised.

Consider some consequences of this lack of integration:

  • Inability to do performance analysis according to dimensions in internal marketing systems like customer segments, high-value vs low-value customers, demographic slices, etc.  
  • Inability to build attribution models focused on customer marketing programs (instead of just acquisition).
  • Inability to capitalise on fractional attribution to inform targeting. While this can be done to an extent to inform digital programs, other tactics like direct mail can’t capitalise on the improved measurement.  
  • Inability to build attribution algorithms by customer segments. The customer segmentation logic typically lives within the marketing database.
  • Difficulty obtaining organisational adoption because multiple measurement solutions are still being used.

To fix the measurement challenges facing marketers today, a single source of truth embedded into a single system must be established. Utilising multiple systems leads to misalignment and lack of agreement. Bridging the gap by integrating all three pieces – top-down, bottom-up, and the customer – is the path marketers must strive toward with the next generation marketing measurement platform.

Some organisations have realised this need for a single platform and are already well down this path. This requires the ability to measure all above and below the line media, external and seasonality factors, and the ability to drill down into the lowest level of granularity (like search keyword), while being integrated directly into the customer marketing platforms to create the single source of measurement truth.

While this may sound like a daunting challenge, there are ways to start establishing the foundation by integrating existing measurement approaches into your customer marketing database. Once the end-to-end process is established, sophistication can evolve over time. An example of a fundamental shift is organisations changing how they code attribution in their marketing infrastructures. By far the most common way attribution is coded in a marketing database today is by assigning a new customer to a media, or source code, or origination code that indicates the one and only source that new customer came in through. 

Long term value

This is convenient from a data structure and reporting perspective, but completely inaccurate. Unfortunately, marketing databases end up being based on just that logic, credit given to one and only one touchpoint. The data structure must change to allow multiple touchpoints to be given partial credit for a sale. Then total credit to a program or media is achieved by summing the partial credit across all sales. Making these types of changes now will enable your organisation to take steps toward realising the long-term value of true connected attribution. For example, the system can allow for adjustment factor coding in a simplified approach at first; analysis, testing, or any other business rules can be used as the starting point. Having the structure in place to encode these additional adjustment factors will enable the reporting systems to further fine tune attribution.  

While progress has been made to unify marketing measurement, the industry still faces significant challenges. It is not uncommon for advertisers to switch measurement vendors every few years because they are not seeing the value promised. A key reason measurement adoption is still a challenge is because the systems are not embedded into the organisation’s marketing infrastructure. Integrating the measurement solution with the customer data within the organisation’s customer marketing database is the fundamental missing piece to establishing a single source of truth.  

Scott Nuernberger

Scott Nuernberger

Scott has more than 10 years of experience developing and implementing analytical solutions into marketing programs for many different companies, including GEICO, AEGON, MBNA, Fidelity, DirectTV, Eli Lilly, GSK, and Dell. Prior to joining Merkle, Scott worked for American Express as a statistician and modeler and taught graduate students statistical methods and experimental design at Cornell University. Scott has dual BS degrees in Brain and Cognitive Sciences and Statistics from The University of Rochester, a MS degree in Statistics from Cornell University, and an MBA from Johns Hopkins University.

Read more from Scott

Join over 10,000 performance marketers for the ultimate weekly update on industry news