No direction home?
Editor's note: Ove Haxthausen is managing director of The Westside Group, a Stamford, Conn., consulting firm.
A specter is haunting market research – the specter of disruption. Sound familiar? I am paraphrasing Marx and Engels’ introduction to their Communist Manifesto because just like communism was going to wreak havoc in Europe for a long time after the publication of that book, market research is currently seeing strong trends of change, some of which could be very disruptive to the industry.
The purpose of this article is not to be a manifesto for one disruption or another but rather to take a look at the different trends out there and then introduce a framework for market research stakeholders – marketers, established market research firms and potential disruptors – to chart a course in the midst of the turbulent waters.
A wider choice
First, let’s take a look at what is happening in the industry. Things are changing in market research, particularly around the collection of data. New approaches have appeared, from biometrics such as eye-tracking and brain activity measurements derived from neuroscience to virtual reality experiments, social media-enabled qualitative market research, social media listening, “quantification” of social media through sentiment analysis and other types of text analytics, big data generated by corporations and other entities, etc. This has created a wider choice of potential data sources for generating market insights.
As corporations and social media generate more and more data just by virtue of the way they operate, there is even a sense that data will be available to measure almost anything and that the need for traditional quantitative and qualitative research will diminish drastically. The value of market research – the logic then goes – will be in the combination and analysis of all this data to extract valuable business insights, rather than in the data-gathering through surveys and qualitative techniques.
Yet, while most market research firms would agree that they deliver value by producing insights, not by gathering data, their business models are not data-agnostic: they are based on fielding market research studies to collect data. Furthermore, the data analysis field is a well-established and crowded one where a plethora of management consulting firms, advanced analytics shops and IT companies already play.
Complement or improve on
In reality, not all trends point toward the irrelevance of market research firms and their business models. Some new data collection techniques are sustaining innovations, the term Clayton Christensen uses in his book The Innovator’s Dilemma to refer to innovations that complement or improve on existing technologies rather than creating a disruption with a radically different technology or approach.
Sustaining new data collection approaches include biometrics and neuroscience brain activity measurement because those are being used as complements to traditional research approaches to provide deeper insights. They cannot replace existing research approaches. Virtual reality experiments are also a sustaining data collection innovation, because behind the virtual reality environment, data is still collected and analyzed in ways consistent with traditional approaches. Social media-enabled qualitative research, social media recruiting of online panels are sustaining, because traditional qualitative and quantitative research is still being conducted using moderators and surveys, even if social media is being leveraged to facilitate the process. Not surprisingly, these sustaining data-gathering innovations tend to be embraced by traditional market research firms because they fit into their business model rather than disrupt it.
Other new data gathering approaches are disruptive to market research firms and their business models. The growing amount of “passive” data is one example. By passive data I mean data that is being generated without any active survey process, where respondents actively have to respond to a questionnaire of some sort. Passive data includes retailer transaction data, Web behavior, recorded customer service calls, social media postings, store traffic recordings, etc.
When passive data can replace traditional “active” data-gathering approaches it becomes disruptive to the market research industry because traditional market research is no longer needed. Text analytics tools such as sentiment analysis play an important role in transforming passive data into information that is comparable to traditional market research findings.
As an example, a brand tracking dashboard by a sentiment analysis provider such as Collective Intellect – now part of Oracle – will contain metrics very similar to what a traditional quantitative brand-tracking survey will provide. The difference is that the sentiment analysis dashboard provides these metrics in real time, continuously. On the other hand, a dashboard using traditional quantitative survey data will only refresh with certain intervals (e.g., quarterly) and with a delay of at least a few weeks to account for data processing.
Of course, the sentiment analysis data will be less complete and likely less accurate: It can’t be mined as easily as the survey data (identification of respondent gender or age, for instance, is hard), the automatic sentiment analysis algorithms may sometimes misinterpret a sentence, respondents may not be representative of a particular target audience and, depending on the volume of social media conversations about a particular brand, some topics may not be available. But the sentiment analysis data comes without any of the respondent engagement issues typically associated with quantitative surveys and is available real time.
Usually faster
Just like disruption theory suggests, sentiment analysis and other tools for analyzing passive data tend to provide lower-quality, less in-depth findings than traditional market research approaches, but usually faster – often in real time, in fact – and at a lower cost.
We can plot the different available research approaches, traditional and disruptive, on a two-dimensional grid where the x-axis represents depth and accuracy of the resulting insights and the y-axis represents the speed and frequency of the insights delivered.
As illustrated in Figure 1, sustaining innovations are placed to the right of traditional research approaches because sustaining innovations tend to increase accuracy and depth of insight. Biometrics and neurological approaches for instance are often used as a complement to traditional approaches to gain further emotional insights. Similarly, social media-facilitated qualitative research is intended to provide an ongoing, more natural forum where respondents can feel more at ease, increasing the accuracy of their responses. As previously noted, disruptive research approaches such as the analysis of passive data tend to provide faster and more frequent insights because these approaches are often fully automated and do not rely on conducting surveys. But as previously discussed, the insights tend to be more shallow and perhaps less accurate than what one would get using traditional approaches and sustaining innovations.
In the bottom-left of the grid we find traditional approaches that have been disrupted by passive data analysis because they do not provide any additional depth or accuracy of analysis that would justifying their slower or less-frequent delivery of insights relative the analysis of passive data. At the top right would be the Holy Grail: 20/20 insight delivered in real time, all the time. But we are not quite there yet.
As marketers weigh their research options as laid out in Figure 1, they have to consider the nature of the business issue they are looking to address. Simply put, are they better off with the 70-80 percent disruptive solution right now or should they instead wait a month or two for the good-old 100 percent solution? It will depend on the issue they are trying to address.
Some business issues are more time-critical than others and some require deeper insights but can wait a bit. As an example, Figure 2 looks at marketing issues typically faced by a consumer goods company, dividing them between those that are very time-critical, where a directional answer now is likely to be the most valuable, and those that tend to require more in-depth insight, even if that means waiting a bit longer for it.
Gain an edge
Not surprisingly, time-critical issues are more tactical in nature but that does not mean they are less important. With the increasing ability to develop insights on a real-time or quasi-real-time basis, this is an area where companies can gain an edge if they acquire that ability to mine data for immediate insights and are then able to act on that insight very quickly for business optimization. For a typical consumer goods company, some of these tactical issues may include in-market product launch monitoring and optimization, to course-correct launch tactics such as promotions, in-store events or advertising in real time depending on in-market performance.
Time-critical issues also include simple business and brand performance monitoring. This is not about solving a specific issue but about ensuring that the business is on-track and being able to investigate and take corrective action as soon as there is a sign of underperformance. For these time-critical issues, the new disruptive research approaches at the top of our grid on Figure 1 are likely to be the most attractive.
On the other side of the time-sensitivity/depth of insight spectrum, we have the more strategic marketing issues. They almost match the tactical issues topic by topic, as illustrated on Figure 2. While the more tactical product launch monitoring and in-market optimization requires time-critical insights, even if only directional, the more strategic new product development work that would have been done prior to launch is less time-critical but requires deeper insight to ensure that the new products will address a compelling need in a new way and provide an edge over the competition. For these more strategic issues, research approaches to the right in our grid on Figure 1 are most likely to work, because they will deliver the deepest, most accurate insight.
Look more carefully
So what does this mean for marketers? Because they have an increasing set of market research options, they need to look more carefully at the issues they are trying to address and at how critical timeliness and depth of insights are to addressing the issue. By categorizing their issues as either time-critical – where depth of insight is less important – or insights-depth-critical – where timeliness is less important – marketers can identify the right research approaches for each issue, looking at whether existing research programs are adequate or not.
This bottom-up approach to determining market research needs will identify what new initiatives are needed as well as what existing programs no longer deliver the right combination of timeliness and depth of insight required for the company to be competitive.
Companies with lots of passive data do this as part of their operations: Since data is free, they undertake data analysis on an as-needed basis to address business issues. Internet retailers such as Amazon and eBay come to mind as best-in-class in this area. Companies with little passive data – such as consumer goods manufacturers, one step removed from the consumer transaction data controlled by the retailers – have to rely more on costly third-party data sources to address their business issues. As a result they can find themselves in a situation where existing research programs and data sources dictate the approach to solving business issues rather than the other way around. Not an optimal situation in a rapid changing data and analytics landscape.
Established market research firms will need to take a look at their offering to assess to what extent it is competitive and identify approaches in the lower-left corner of Figure 1 that are candidates for disruption by passive data analysis because they offer only limited insights depth, in a less-frequent or timely manner than passive data analysis can and perhaps even at a higher price.
Likely to be hard
Continuous tracking surveys focused on brand performance or consumption may be particularly at risk unless they can also deliver sufficient depth of insight to serve as a basis for strategic issues such as brand positioning or new product development. At the same time, market research firms may explore ways to incorporate passive data analysis into their offerings. If recent events are any indication, this is likely to be hard, because established research firms correctly see these approaches as potentially disruptive to their core businesses. As an example, Cymfony, a text analytics firm previously part of WPP’s Kantar, was not integrated into any of Kantar’s traditional market research firms but rather had its own competing sales force. In April last year Cymfony was bought by Visible Technologies, a social media monitoring company in which WPP also had a stake. The transaction did not reduce WPP’s stake in the two companies but it removed Cymfony from the Kantar fold.
Of course the risk established research firms run by ignoring passive data analytics is that they will no longer be able to address the time-critical tactical marketing issues because of the disruptive power of passive data analytics. They will then have to make sure that their offerings offer deep enough insights to address the less time-sensitive strategic marketing issues.
Finally, an emerging passive data analytics firm wishing to play in market research will have to acquire more marketing expertise and access to the marketing community. These firms’ skills are in data science and analytics as well as information technology but they often have less functional expertise in areas such as marketing. This makes them an obvious target for an established market research firm willing to take a risk with them but more likely they could find a match with a marketing or branding consultancy interested in venturing into addressing the more time-critical tactical issues as well. Of course, passive data analytics firms could also build their marketing expertise in-house.
Decide where they want to play
In summary, no one-size-fits-all “killer app” is likely to emerge anytime soon from the increasing variety of market research approaches. But the emergence of new approaches, particularly the disruptive potential of passive data analytics, makes it imperative for established market research players to review their offerings and decide where they want to play. Because of the increasing set of options available to them, marketers will have to focus more on the optimal timeliness vs. insights depth trade-off required for the issues they are looking to solve. And emerging passive analytics firms will have to acquire more functional expertise in marketing in order to better compete in the field.
Navigating turbulent waters can be either exciting or scary, depending on how prepared you are and whether you have set a clear course. That is your choice.