5 Tips to Understanding your Consumer Tracking Data

Often times when analysts begin working with a market research data set, it can be a daunting experience to know how to get the most value from the wealth of information that may be available.  How do I best use the data?  Where is the story? How do I change this analysis from a data dump to meaningful insight? I need to look smarter and not waste people’s time with numbing amounts of statistics.

Here are the 5 best tips we received from our own research of data users:

  1. MIX CURRENT AND TRENDED DATA – Start with the most important high-level metrics and take a current view and trended view of the data. Most analysis starts with looking at the current period of data.  Everyone is familiar with the expression “what have you done for me lately?”. That is why most time is spent on the most recently collected data, but it is helpful to always see where you have come from to help see where your product, brand or service is heading.  If it is available from your research sources, a trend view is always helpful.

    2017 Wireless Carrier Overall Experience (% 9- 10 on 10-point scale)

    Survey 1 – 2017
    Brand 3 is highest rated wireless carriers with steady improvement from 4th to 1st

    Survey 1 – Trended
  2. USE MULTIPLE SOURCES – Many analysts fall into a trap of using a ‘go to’ source. Make sure you are rounding out the analysis with multiple sources: syndicated, proprietary and secondary (e.g., reviews).

    Below the results from the 2nd source are consistent, but the 3rd survey shows a different story.

    Syndicated Survey 2
    Proprietary Survey 3
  3. SELECT KEY BENCHMARKS – Benchmarks might be total sample, a given market (region), or a particular best-in-class product. You should have the flexibility to easily establish your benchmark to compare differences. Use simple significance tests or indexed data to identify statistically relevant data points to discriminate against the benchmarks.

    In the tables below, Brand 3 and 5 are highest in NPS (net promoter score) compared to all other carriers. When compared to Brand 3 all carriers are performing significantly lower on NPS.

    Significant differences of Carriers to Total (All Wireless Providers)

    Significant differences of carries compared to Brand 3 (highest NPS)

  4. IDENTIFY TREND DIFFERENCE – For trended data you can find your story by looking at differences in various ways. For example, below there is a consistent story of service and reputation when selecting Brand 1 against the competition, while wireless connectivity and the variety of service plans are weaker reasons to select Brand 1 against their competitors.

    Period to period significance is important when looking at tracking studies to measure where there is improvement or decline. Most analysts need to measure brand health by answering where are we better, worse and unchanged. Brand 1 is showing increased period to period scoring (2015 better than 2014, 2016 better than 2015, etc) on its two main consumer drivers (service and reputation) the last three periods.

  5. SHARE FINDINGS – DEMOCRATIZE THROUGH THE ORGANIZATION – Most repeated analysis from tracking studies should be shared throughout the team and organization. The ability of a non-analyst to navigate findings through a visual dashboard interface has been growing in importance, and these non-analysts should be able to intuitively share updates, insights, and stories.

    Sharing of findings might require a different deliverable to different audiences. KPI (key performance indicator) dashboards can be used from C-Suite to the lower levels of the organization. Guided reporting dashboards are usually required for users that might need more frequent interaction with market research findings.

    Unlike your heavier duty market research tools (SPSS, SAS, etc), an ideal research dashboard provides the audience with the ability to simply visualize the data with:

    • Key banners and / or filters
    • Current period / wave or a trended view
    • Option to export to PowerPoint or Excel
    • Built in sample size thresholds and / or significance testing
    • Ability to save page views to build a ‘Storyboard’

For more information about this blog content, contact John Sevec (jsevec@mtab.com)