As companies struggle to extract value from brand tracking, many are turning to new methods to understand consumers, integrating new data sources and using new technologies. Following her session, “Innovations in Brand Tracking” at The Quirk’s Event in Brooklyn, N.Y., in February, Quirk’s interviewed Suzanne Henricksen, senior director of global insights and consumer affairs at The Clorox Company, to discuss the decision to re-create brand tracking at Clorox. From the first steps of choosing methodologies to the challenge of providing real-time data, Henricksen provides information on Clorox’s journey, along with tips for marketing researchers looking to get more out of brand tracking.
Mike Clarke, executive vice president, general manager at Lieberman Research Worldwide, co-presented the Quirk’s Event session with Henricksen.
Your session provided a detailed look at the decision to challenge the status quo and re-create brand tracking at Clorox. Could you describe the process and the approach you took?
The process was to first step back and realize the world had changed – both marketing and research – but our brand tracking had not evolved. So we embarked on an extensive exploration to see what new ideas were available. After some informal exploration, we started an RFP process where we posed this challenge to potential partners and actively encouraged different thinking. We explored a spectrum of methods – traditional, social-only and options in between. We did pilot waves with four different partners across multiple brands and, based on the results from each pilot, we decided to proceed with an approach integrating work from two leading-edge research partners … one more traditional/asking and one more social/listening in focus.
What tips would you have for researchers looking to incorporate innovative approaches with more traditional methods?
First, get a commitment from your business partners that change is needed and implementing the best approach is more important than trending results to the past. Despite setting this expectation up front, you need to work behind the scenes to be able to make some comparisons to past results, as the question is unavoidable even if everyone is committed to change at the beginning.
It is also important to set expectations with business partners and your team that you will all need to invest more time at first due to new, more sophisticated approaches.
What new techniques or technologies did you decide to leverage when looking at brand loyalty and loyalty drivers?
We used Bayesian network modeling. Compared to basic driver analysis, this approach helps us better understand how various associations drive loyalty and how they impact each other. Basic drivers tell us the impact of quality perceptions on loyalty. Bayesian networks also tell us the impact of quality on loyalty plus the impact of quality on value and all other attributes in the model. This provides several types of insights you don’t get with basic drivers.
We can identify alternative paths to loyalty based on our organization’s own strength and weakness. Sometimes top loyalty drivers are not levers you feel your brand can or should deliver, so you need the insights to develop a meaningful Plan B.
With a best combination of attributes to drive loyalty, we can do a TURF-like driver analysis which determines combinations of messages that best drive loyalty vs. the individual rank order of attributes in isolation.
What research methods have you found to be most successful in gathering consumers’ less-conscious perceptions of brands?
As researchers, we need to seek out solutions that are pragmatic, more real-time and offer more sophisticated ways to understand these dynamics, but also ones that are easy to understand and agile to execute.
We have tried-and-true tools that do a good job of measuring rational, conscious dynamics but these fail to deliver that deeper understanding and are not grounded in the science of emotional and non-conscious measurement. We also have tools to go deeper into emotions and non-conscious decisions but they can often be costly and time-intensive. In addition to those challenges, the insights are interesting but not always relevant to taking action.
In the Clorox brand-tracking project, one more pragmatic technique we used was a rapid choice exercise. This approach uses speed of association – between a brand and word – to understand gut, less-conscious brand perceptions. On the real-time, voice-of-the-consumer front we used a social listening approach with algorithms designed to filter out the noise and home in on the impact of these authentic consumer conversations on brand health. By bringing “asking” and “listening” approaches like these together we get a more comprehensive, accurate, timely and relevant point of view on the consumer and the health of our brand leading us to make better, more impactful decisions.
In your session you mentioned the importance of being “data agnostic.” Describe what that looks like and why it’s so essential.
Being data agnostic is in reference to the types of data we use to answer business questions. Traditionally, market research has been conducted largely with online surveys, focus groups, etc. We now need to seek out additional data sources to augment those traditional sources – social listening, big data analysis ... We need to define our core job as using consumer data to answer business questions, using any and all data available. We should not define and restrict our understanding of a consumer behavior or insight based on singular data sources or specific types of data that we maybe once believed were the only way to get an answer.
Many researchers turn to social platforms to get a deeper understanding of consumers. Could you describe some common mistakes that brands make when integrating social data? How can these be overcome?
Don’t overreact and don’t underreact.
Social listening is new and exciting and the data is more dynamic, so it’s a common mistake to overreact by jumping at any spike in the conversation. You need to develop expertise in how you interpret results and also triangulate answers with additional points of data.
We also make the mistake of not taking action. At this early stage in its adoption into market research, too many people discount the results because it’s not the representative sample that traditional methods provide. Anyone talking about your brand is an important consumer to understand. Don’t dismiss that conversation, just make sure to interpret it correctly.
Providing real-time data is essential to staying ahead in today’s market. What steps would you recommend to researchers looking to avoid common research pitfalls caused by time constraints?
The market is moving faster, so we need to be more responsive – which means we need early indicators and more agile tools. Options include social listening, customer service center data, real-time data collection and reporting of online survey or social results through portals and communities.
What do you see for the future of marketing research in regards to brand-tracking trends?
Many of the themes already discussed are absolutely key to the future of brand tracking. In particular, emotional/non-conscious measurement, incorporating real-time listening data that truly represents the unbiased voice of our consumer and having early indicators/more agile tools. All these will be key to understanding our more complex and rapidly changing environment. Also, having approaches that have strong financial linkages, correlating to in-market results, will continue to be critical.