This is the third part of our ongoing series to understand some of the dynamics that explain how companies in the Walker Index outperform the market by more than six-to-one. In the first two segments, we examined the roles that Relevance and Alignment and Team and Resources play in world-class customer listening and how the Walker Index companies perform on these dimensions. In this entry, we will focus our attention on the area that is most closely associated with customer listening – the Information Gathering process.
Let’s start with a strategic view of the Information Gathering process. Why is this step important? To quote our client, IHS CEO Jerre Stead, “Facts are our friends.” Given the right data, we can make very informed, strategic decisions. For this to be the case, however, there are some conditions that have to be met:
1) We have to actually gather and use the information – This sounds horribly obvious, but I have seen examples in my career where companies invest considerable funds to gather data only to use a single number (or discount the results altogether).
Moreover, in companies that have an abundance of data, the company can suffer from “analysis paralysis” – that is, they have so much data it becomes debilitating – or, worse, they can be so silo-oriented that they do not know what they have across the entire enterprise, which means they either gather the same data again or make decisions on gut feel alone.
2) We have to know who the feedback represents – For us to be able to generalize the results of the listening program, we first have to know what the results are representative of. From the perspective of a traditional customer listening program, this means that customer lists have to be complete (no cherry-picking), up-to-date (particularly if you are segmenting by product usage, geographies, etc.), and (if you are using a sampling approach) randomly selected.
This has relevance at the account level as well – as we focus more on taking action and linking results to financial behavior, our core unit of measure is often the company, not the contact within the company. Therefore, it is important that we have a diverse set of informants within a single account so account leaders can have a comprehensive view of the health of the entire account.
3) It is not just about quantitative data – Statistical models are nice, but, to quote George Box, “all models are wrong, but some are useful.” This essentially means that any statistical model has limitations, so it is important to be able to evaluate the limitations (via R-squared values, for example) and provide additional context. This is where unstructured data come in.
“Unstructured data” is what we used to refer to as verbatim comments, open ends, write-in comments, etc. Unstructured data is important in a number of ways – it can provide context/flavor to the quantitative numbers, it can uncover issues/factors not accounted for in our model, and, increasingly, text analytics are allowing us to converting the unstructured data into concepts that can be incorporated in modeling.
4) We need to be mindful of other listening posts – I touched on this above; before you start gathering new data, be sure you are fully leveraging the information you have already gathered – it may be good enough to answer your business question.
We counsel our clients to warehouse their customer listening data in a common database application; this makes all the data accessible from within a single platform, it allows us to control how often a customer is solicited for opinion (to minimize customer burnout and keep sound response rates), and it enables us to see how a single customer’s sentiment changes over time (this will be very important when we discuss the step of Validation).
This is also where social media fits in. I have seen three schools of thought emerge as it relates to social media:
a) Social media will replace traditional research entirely;
b) Social media has too many limitations to be a viable decision-making tool;
c) Social media can be used to complement traditional listening tools/approaches;
I tend to agree with the more moderate third approach – clearly, social media has some technical issues (it likely does not represent the entire customer base, the signal-to-noise ratio is lower than we might like, etc.), but it does provide additional context and insight that can help provide clarity on any issues that may exist within the customer experience.
Among the Walker Index companies, there is strong performance on the topics of having structured listening programs with sound sampling/customer list management practices and incorporate unstructured data. Perhaps not surprisingly, the assimilation of unsolicited data (such as social media data) into the results has a low adoption rate at this time. This, however, is changing – as companies take better steps to integrate their data (through CRM installations, for example), the ability to leverage these data increases. It is an area that will continue to improve into the near future.
Many customer listening programs begin – and, unfortunately, end – with the Information Gathering process; therefore, this is probably the area where most programs have alignment. As we know, though, it is not sufficient to just gather the information. In part four of this series, we will focus on the importance and benefits of Communications in the world-class customer listening program.
Mark A. Ratekin
Senior Vice President, Consulting Services