If you work in marketing research or brand management, you'll likely have come across brand tracking studies (trackers). At first glance, trackers seem to be a relatively straightforward methodology for assessing brand health. That is until they start to “come off the rails,” and that’s when things get interesting. In this post, I’ll cover the three main considerations for conducting successful trackers.
Sample representativeness
When starting a new tracking program or taking over an existing one, I usually recommend conducting an initial data collection wave to determine the appropriate sample characteristics and proportions. If I’m surveying consumers, I use data from Census or Pew Research Center to inform targeting; if it’s a B2B study, I use The Bureau of Labor Statistics.
I've often seen legacy sample group proportions or data weighting that no longer represent the audiences of interest. Depending on the category, the target groups and their proportions should be assessed and adjusted as needed over time. Bottom line: segments change. Is your sample still representative of the target audience you're tracking?
Consideration 1: The keyword above is “inform.” More often than not, the final sample composition will differ from the completed interviews due to screening questions, first response biases, over-quotas, survey characteristics, etc. When developing the proportions or weighting scheme, it’s vital to examine profiles of the “first clicks” to the survey vs. completed surveys. The first click tells me how representative the sample looks before screening, self-selecting into the survey, and over-quotas.
Upon determining target sub-groups and their proportions, it's essential to stratify the recruitment and apply as minimal quotas as possible to ensure consistent sample composition and comparability in subsequent waves.
Survey instrument experience & accuracy
Brand tracking studies have been around a long time, and many have evolved from the early days of the Internet or even longer, from telephone studies.
Consideration 2: Choose the appropriate mode of data collection and optimize the user experience. Your online survey must be concise and user-friendly on both a PC and mobile.
Residential phone services are becoming extinct. In 2017, expenditures on landlines were 17.6%, according to the Bureau of Labor Statistics; cellular expenditures were 82.4%. What do you think it is now? According to Pew Research, 97% of adults owned a cell phone in 2021, most of which were smartphones (85%).
Why is this important? In the mid-2000s, when I first started tracking devices used, only about 12% of respondents accessed my surveys on the phone. These days, that’s north of 60%. Yet many surveys are not optimized for smartphones usage.
The following examples are survey questions on a mobile phone I’ve seen recently (And yes, I look for survey design innovations during my free time. LOL). How would you or could you answer these questions. The one on the left looks like a telephone survey that morphed online.
Too long and disjointed
It’s generally accepted in the market research industry that an online survey should be no longer than 20 minutes at most. From my experience, I believe the optimal length is 12-14 minutes, which represents a good balance between time to complete, respondent attention span, and a reasonable incentive amount.
Consideration 3: Focus on the questions that drive brand health equity. I strongly suggest conducting a key drivers analysis to identify the most influential attributes at various points in time but especially at the start. Are you monitoring the right attributes? Avoid making the tracker into a Franken-zero by pasting on “nice to know” questions that increase survey length and lead to a disjointed respondent experience. Instead, make it a Franken-hero by designing the tracker in modules, which contains the brand health KPIs flow with the ability to add current, important topics that come up but don't have to be tracked ongoingly.
Ultimately, the more protracted and disjointed the survey, the more noise in the data. Vet any potential additional questions rigorously, i.e., does it need to be tracked, or is the burning question of the week more fitting elsewhere? Audit and pressure test your instrument regularly to ensure that the most critical issues are addressed.
If you must add questions, put them after the main brand KPIs so they don’t disrupt flow and context, thus influencing responses to subsequent questions. For example, after the global warming and baby seal question, I can already hear another question from the CEO: "why did we drop significantly in our brand's sustainability rating?"
Well there you have it! My three main recommendations for how to keep your brand tracking study healthy. If you agree or disagree or have other recommendation, let me know. As always, I’m happy to discuss over lunch or coffee on me.
Email me at rsilvestre@strategence-us.com.