FEATURE29 April 2015

Is it time to rethink the tracking study?


In the first of his series of features ‘rethinking’ some of market research’s sacred cows, Colin Strong turns the spotlight on the tracking study.

Continuous survey-based tracking programmes have, for many years, been the cornerstones for most brands’ measurement and analysis of their markets and customers. 

Done well, they are an essential ingredient for performance monitoring and strategy development. And they often represent the bulk of revenues for many market research agencies.  

Many have pontificated about the future of the tracker – much of the discussion has been about its impending death, which has proved to be premature. Of course tracking studies often have their downsides. Tracking programme logistics can mean a disconnect between availability of relevant and timely outputs and the needs of the business. 

Because of the nature of the sampling typically (and necessarily) employed, there is often not an opportunity for longitudinal analysis and therefore more robust ROI analysis. Sampling designs of trackers mean it is often difficult to integrate additional data sources at an individual respondent level. 

Of course, no one research approach is perfect and there has been a great deal of creative thinking applied to tracking studies to design solutions to help mitigate some of these challenges.  And a very broad range of solutions sits under ‘trackers’.

Nevertheless, a number of themes are driving an agenda of change in tracking including:

  • the business environment is becoming more agile, requiring speedy insights for rapid decision making
  • increased availability of consumer data is changing expectations of what we need to collect from tracking surveys
  • company margins are being squeezed affecting tracker budgets.

With these points in mind I have identified key themes that I believe will start disrupting how trackers are conducted:

Trackers as data integration tools: It is not enough for tracking studies to sit outside of the other data sources that are increasingly available, either from within the organisation (e.g. their own transaction data) or from other data sources. For example, midata initiatives already allow access to data such as banking activity. We will start to see much more activity to bring sources together and draw insight from them.

Embracing the longitudinal: There is an essential dilemma with tracking studies – we cannot ask the same people the same questions month after month as this will affect their responses, but we also want to examine the way in which attitudes and behaviours change over time – at a respondent level. Marketing technology is changing expectations of analysis to drive activity and measure ROI. We will see new techniques emerging to use research communities alongside independent sampling to find creative middle paths through these issues.

Understanding the context: As market researchers we tend to use the individual as our unit of analysis, not least because of the tools historically available to us.  But there is evidence that it is as much the social context in which we find ourselves that drives our decision making rather than our individual needs and attitudes.  In the moment research, life-logging tools, social media and a range of other data sources will switch focus to measure the impact of context (and not just our responses) on consumer behaviour. 

Human insights from raw data: There is an avalanche of studies that show the way in which we can use a variety of data sources to derive new understandings about the inner workings of our minds. These insights are often time-consuming or even impossible to derive from direct questioning. As we get better at this, we will increasingly use these new measures alongside attitudinal data. 

The re-emergence of interpretive research: One of the areas in which new data sources will not serve us well is the meaning that consumers derive from their experiences. We may see what people do (and indeed ask about this less), we may be able to predict what people do and some of the time we can work out why. But given we are not stimulus-response creatures, we attach our own personal meaning to the process which influences, or even determines, outcomes. The questions we ask will be reframed with this in mind and we will see a growing interest in applying interpretive approaches to the quantitative field.

Lean but not mean: The fact that tracking studies can become large and unwieldy is both a marker of their perceived value but also a symbol of how they are not always used strategically. There will be even more activity to embed the tracker study within a broader programme of consumer engagement, such as a research community. This means that the core questionnaire can be kept to a minimum while less strategic or tactical issues can be rapidly (and cost effectively) addressed by other means. 

Tracking studies often have negative associations for many in the research industry. I can see why, as when done poorly they can be expensive, slow and out-of-touch with the company’s strategy. But in my experience, when done well, they can be an excellent means of performance monitoring and strategic analysis. 

My sense is that we are coming to a point where even those trackers that are good examples of survey-led insight need to be reinvented. Marketing technology is booming, sentiment is with this rather than market research because of the promise of what it can deliver. The research industry can, and should, grab the initiative back but to do so means rethinking some of the research approaches that we once took for granted. 

Colin Strong a board member of Verve. His book Humanizing Big Data is available from the Kogan Page website, where readers can gain a 20% discount by using the code MKTHBDS .  

1 Comment

5 years ago

I'd also like to revisit trackers for some basic methodological reasons. We are much smarter about what makes a good question and a good questionnaire. Given those criteria (simple human language, shorter), most trackers need a complete re-write of every question and answer. However, we are often unwilling to do so because this will "change our norms." This is the most short-sighted (stupid?) reason to not fix a tracker. Because we don't want our bad data to change. Revisit your tracker now. Fix it now. Restart your norms now so that you only have 3 years of bad norms not 8 years of bad norms.

Like Report