fbpx

Seeking Clarity on Campaign Measurement

  • Earl Bless – Researcher, Brand and Marketing Impact, Intel Corporation Christina Flint – Researcher, Digital Marketing and Media, Intel Corporation Edwin Derks – Sr. Marketing Analytics, Intel Corporation

Summary

The Intel team presented learning from Intel’s 2017 research on campaign measurement methodologies. These insights are from a yearlong “research on research” study to determine the strengths and weaknesses of campaign impact.

The presenters provided background on Intel, explaining that although the company is known for microprocessor chips, the company is also involved in autonomous driving, artificial intelligence, etc.

There is not one best method for campaign measurement. Among the characteristics of a best method would be: consistency; eliminates or minimized bias; measured cross-channel impact; ability to optimize media.

Three research methodologies used to test two 2017 campaigns involving two TV bursts + digital media: Passive (TV and digital); TV OTS (opportunity to see) and passive digital; stated recognition.

Findings:

  • Stated recall and OTS shared similar end results for campaign KPIs, including significance and overall lift.
  • Significant levels of stated recognition within control in Passive and OTS methods were observed.
  • “Ghost” recognition exists in stated measures. That amount was consistent regardless of stimulus.
    (Intel has been using the same spokesperson, Jim Parsons, for a while and this might have impacted measurement.)
  • At this time, stated measurement, with the addition of an efficiency metric, provides the best overall measure of campaign, cross media impact and campaign efficiency.
  • Intel is using cost per person impacted (CPPI) to gauge campaign efficiency across channels and campaigns.

Intel’s plans moving forward:

  • Use passive measurement for digital only campaigns that need in-flight optimization, or to answer questions that cannot be answered by stated recall.
  • OTS for certain offline media e.g. TV or OOH which limits its application if used alone.
  • Stated awareness for most campaigns. Add CPPI to provide comparable efficiency measures.

Key takeaways:

  • There is no single best method for campaign measurement. There are pros and cons to each approach. Speed and granularity differ for various measurement methods.
  • Data contamination is a problem.
  • “How are we doing and are we doing better?” are the questions asked by Intel’s management. To measure efficiency, Intel began using cost per person impacted (CPPI) in order to answer these questions and to show lift in KPI, stated recognition as well as CPPI. CPPI allows comparisons between channels and between campaigns. CPPI can also be used for optimization purposes.

Presentation:

 DOWNLOAD PDF

Presented at AUDIENCExSCIENCE, June 12-13, 2018

Add a Comment