Voice of the Customer (VoC) data has come to represent a very broad spectrum of data streams, from traditional consumer research, convenience surveys, comment cards, call center verbatim and all the way through to social media. Too often, similar expectations are placed on these data streams even though the nature of the information varies enormously. Many times the data is positioned as similar or in competition, and businesses awkwardly apply the same analytic tools and reporting structures on data streams that are wholly divergent. But each data stream offers a unique type of feedback, which should be appropriately applied to the business.
To help organize your data and manage stakeholders expectations, it is a useful exercise to plot your VoC data along two dimensions – representative to user initiated, and as either structured (focused) or unstructured (open ended). This outlook helps to quickly organize the tools and processes available.
Use Representative Data to Measure Strategy
Of all the different types of data collected, representative data, which is based on random sampling, is the only data that should measure the success and failures of strategies and drive strategic changes. Why? Because representative survey methods give equal opportunity to all opinions and while a true random sample is a theoretical goal, by paying attention to collection design, the more reliable your results will be and the more confident you will be in the strategies you implement.
Focused representative research is important for clear customer insights around KPI’s – such as measuring purchase intent, satisfaction, or brand. This is the type of data where statistics apply and it is from this data that testing, performance tracking, and predictive analytics can be done. From this, programs and strategy can be quantifiably evaluated.
Within representative sampling, if you make the research very open-ended, you can give customers the freedom to add context or opinion. This is exceptionally useful for exploration, understanding needs and desires and performance gaps. Text analytics and pattern recognition tools can help reap the benefits of this data. This approach can serve an invaluable source of new hypotheses that can drive business innovation.
User- initiated Feedback Can Drive Process
User-initiated feedback – such as comment cards are very different than representative feedback, as it is not random. Research shows that users who are motivated to provide feedback unprompted generate data that is heavily weighted towards the negative. If companies use this data to make strategic changes, the unbalanced nature of a non-representative sample can cause companies to make poor decisions based on the misguided zeal to please a few, vocal users.
Although this type of data does not serve the same function as representative sampling, its temporal nature demands quick follow-up, and is an invaluable source for individualized information with which the business can save, convert or nurture vocal customers. This type of feedback may generate volumes of data and have the trappings of representative data, but it should be viewed as an advisory service, with an opportunity for businesses to actively engage this important user segment and use it to create studies to test broader strategies.
User-initiated feedback that includes open comments or verbatim data provides the largest net from which to capture as much feedback as possible to try and improve the business. It has the unique challenge of searching for relevance among the noise. Identifying specific problem areas lends itself to data mining and pattern recognition techniques. New big data processing and machine learning techniques derived from search have significantly increased a business’ ability to take advantage of this data.
Structured individualized feedback provides more direct focus. The most extreme example is product ratings or likes, which can give a simple comparison and is useful to other users considering the same product or service but this can only provide so much insight. Focused feedback through simple links or forms can provide targeted feedback for the business and should be applied to specific applications or processes to categorize issues or needs. Understanding specific issues in the checkout procedure is an example of exceptionally tactical bug fixing along with a clear opportunity for sales management.
Keep an Eye on Social Data for Future Use
Social data is an interesting beast. It is by definition individualized unstructured feedback. Managing support issues, fostering loyalty, and nurturing advocacy are key use cases from social data. Because social data occurs in such volume and its penetration is growing, many are trying to figure out whether it can be leveraged as market research. For now, it is best to keep representative tracking in place for those big decisions and important KPI’s, while beginning work on contrasting results from social with standard methods so you can decide what elements can be absorbed. While social may prove very useful on the unstructured side of representative, it will take more effort to replace the testing and monitoring side.
Embrace Multiple Data Streams for a Competitive Advantage
Leveraging the right data for the right issue will ensure solid decisions are made. Strategic decisions based on a skewed sample will not help the business grow, for this you want a representative sample. At the same time, you cannot afford to ignore a clear client problem that is costing you revenue opportunities or customer retention. It is imperative to build in many points of contact to collect the voice of your customer from surveys to comment cards and take the opportunity to know your customers better than your competition. Also, the more aligned you are with when your customers want to give feedback, the more relevant your data will be. So embrace multiple data streams and don’t make them compete for attention.
First published in Destination CRM
Image source: Alan Levine