Editor's note: Gerry Katz is executive vice president of Applied Marketing Science Inc., a Waltham, Mass., research firm. An earlier version of this article appeared in the October 2013 issue of Visions, published by the Product Development and Management Association, Chicago.
In 2006, I published an article entitled “Hijacking the voice of the customer,”1 in which I condemned the misuse of the term voice of the customer (VOC) and those who were unfairly criticizing its effectiveness. The way they talked about it felt like a hijacking – forcing it to a destination where it was never intended to go! Well, it’s still happening, although today, it’s not all bad. In fact, some of it is actually welcome.
I observe three major viewpoints in the field of VOC, each with a group of staunch advocates leading the way. I’ve dubbed these exemplars the Naysayers, the Stretchers and the Technovangelists.
Before we explore each of these groups, let’s go back to the beginning. The roots of the term voice of the customer come from a Japanese product development methodology called quality function deployment (QFD). QFD requires a cross-functional team to complete a detailed matrix (referred to as the house of quality) or series of matrices, the purpose of which is to prioritize a large set of potential product development activities. QFD starts with a detailed list of customer needs that make up the rows of the matrix. This list of needs became known as the voice of the customer. The term was institutionalized in a landmark paper of that name by Abbie Griffin and John Hauser and published in 1993 in the journal Marketing Science. This paper, inspired by Griffin’s doctoral dissertation at MIT’s Sloan School of Management, empirically identified best practices companies should use to understand customer needs. The paper won many awards and remains the foundational work in the field more than 20 years later.
Griffin and Hauser define VOC as having four parts: a complete set of customer wants and needs; expressed in the customer’s own words; organized by customers into a structured hierarchy; and prioritized by customers for relative importance and current performance/satisfaction.
In short, VOC is and always has been about the study of customer needs, and was intended for: developing new or improved products; designing new or improved services; and improving business processes.
Given these roots – specifically intended as a foundation for innovation – the term has now morphed quite a bit over the years, which leads us to our three current trends.
This first viewpoint is decidedly unwelcome: VOC simply doesn’t work. The Naysayers argue thusly:
“VOC will never let you be more creative than your most creative customer.”2
“Ask your customer what they want and you are bound to get yesterday’s product...”
“If I’d asked customers what they wanted, they’d have said a faster horse!” (Henry Ford3)
“You should ‘ignore’ the voice of the customer, because no truly innovative idea has ever come from a customer. The customer is the dumbest guy in the room!” (Michael Treacy)
“It’s hard for [customers] to tell you what they want when they’ve never seen anything remotely like it.” (Steve Jobs)
All of these comments reflect a fundamental misinterpretation of the definition of VOC and, more importantly, a naïveté about how VOC research should be conducted. A common amateur mistake, ironically, is to ask the customer, ”What do you want?” When carried out with this kind of direct frontal assault, the customer assumes that they are supposed to describe the exact feature or solution they want. This approach confuses needs with solutions to those needs.
The Naysayers misunderstand the role of market research in innovation. Anyone who thinks that VOC just means “asking customers what they want” is sadly mistaken and any researcher or product developer who has ever tried this approach understands its futility. Why? Because most customers aren’t very good at coming up with innovative solutions. And frankly, that’s not their job – it’s yours!
Great voice-of-the-customer research focuses on what the customer is trying to do and what makes it hard or easy to get it done. Clayton Christensen and his colleagues describe it in particularly colorful terms: “When customers find that they need to get a job done, they ‘hire’ products or services to do the job.”4 He then goes on to make the case that it is better to focus on the job than on the product. Similarly, Griffin and Hauser define a need as “A description, in the customer’s own words, of the benefit to be fulfilled by the product or service.”5 They go on to say that it has nothing to do with the way in which the product or service delivers on that benefit, only with the desired benefit itself. With a combination of observation (ethnography) and probing conversation (lots of “why” questions), the voice-of-the-customer process helps us understand the customer’s needs, both articulated and unarticulated. But from there, it is the innovator’s job to come up with the new features or solutions that respond to those needs. Most customers simply lack the technical understanding, the creativity or both.
Steve Jobs understood this distinction, even if he didn’t articulate it very well, as he went on to say, “It’s not the consumers’ job to know what they want.” In fact, it is their job, usually with some help, to convey their wants and needs – either verbally or non-verbally – in accomplishing jobs or tasks. What is not their job is to come up with the exact solutions.
Whether Jobs did market research or not (Apple insiders disagree on this point), he could innately sense customer needs so that Apple always seemed to be working on the “right” problems. But what really set him apart from other product developers and made him a serial innovator (another term coined by Griffin6) were his solutions – clearly more elegant and intuitive than anything the world had seen. But unlike Jobs, most innovators feel relatively unsure about customer needs. In study after study, it’s still listed as the No. 1 reason for new product failure. So most innovators still need to do market research. There is no guarantee that they will come up with good solutions, only that they’ll be working on the right problems.
Many product developers are still quick to call the words and success of Steve Jobs as the main pieces of evidence in the case against voice-of-the-customer research and my objection is always the same: If your CEO is as intuitively creative and visionary as Steve Jobs, you don’t need anyone’s help. If however your CEO is like the rest of us, you could probably benefit a lot and significantly reduce risk and time-to-market by spending some effort and money in the front end of innovation to research your customers’ needs. You’ll still need to do additional research down the road to test out concepts, evaluate prototypes and fine-tune your messaging but at least you’ll be working on the right problems.
The second viewpoint, while somewhat veering away from the original definition, is actually quite welcome. The Stretchers have broadened the concept of voice-of-the-customer with points like:
“VOC is just another term for market research. Any interaction with customers is VOC.”
“VOC is about staying more in touch with your customers.”
“VOC is for improving the customer experience.”
“VOC is where we share our new product ideas with customers to see if they like them – just to be sure that we’re on the right track.”
The Stretchers take VOC well beyond its original definition. For many, VOC has become a euphemism – it sounds better than “market research” or what we once called customer satisfaction measurement (and its more recent incarnation, the Net Promoter Score) and customer relationship management (CRM). These are good things to do but they are far afield from what Griffin and Hauser were talking about, because most of these new VOC initiatives focus on existing products and existing customers, rather than next-generation products and potential future customers. Furthermore, most have a decided bias toward the customer service and customer experience components rather than on the product itself.
However, this expanded definition of VOC remains true to its origins if the objective is service and/or process improvement. Remember that the goal of VOC is to provide a springboard for innovation, whether it results in better products, better services or better business processes. The major departure is that this new definition places less emphasis on new product development and it typically doesn’t include non-customers, competitors’ customers or ex-customers – important sources of information for the development of new and better products.
So what about these Stretchers – those who use the term VOC to mean customer satisfaction measurement, customer relationship management or customer experience management? Is there still a connection? There certainly should be, although in practice, this connection is often downplayed or ignored altogether. Whether a company uses overall satisfaction or the Net Promoter Score (i.e., the likelihood to recommend), both have become a sort of report card that indicates how well the company is doing in pleasing its customers. Whenever the score rises, there is much self-congratulation and in many companies, additional bonus payments. A falling score, however, invites enormous hand-wringing and finger-pointing (despite the fact that it might just be statistical noise). This is reminiscent of when Johnny brings home an “F” in reading. Many families reflexively blame the teacher or the grading system instead of asking the more fundamental question: “Why can’t Johnny read?” A report card is never prescriptive, it is only descriptive.
Here’s where the old and new definitions connect. In addition to the overall satisfaction and likelihood-to-recommend questions, these VOC-centric customer satisfaction programs ask customers to rate a company or product along one to two dozen individual attributes. Whether implicitly or explicitly, these attributes are assumed to be the drivers of overall satisfaction or likelihood to recommend. And indeed, with many of the more sophisticated systems, there is a formal effort to build an empirical model where overall satisfaction or the Net Promoter Score is used as the dependent variable and the individual attributes are treated as independent variables:
overall satisfaction or Net Promoter Score = ƒ (attr1, attr2, attr3, … attrn)
Although there may be many analytical problems to contend with such as multicolinearity, dealing with discrete, categorical or binary variables, etc., the goal is to come up with a coefficient for each attribute whose relative magnitude is an indicator of how much that attribute is assumed to be driving overall satisfaction or the likelihood to recommend.
Where do these attributes come from? Many companies simply resurrect and tweak an old list from previous research or do some focus groups to come up with a list, with little rigor or deep thought. Here, traditional VOC can make a huge difference, providing a rigorous way to identify those attributes. Why is this important? Because the real drivers of overall satisfaction or likelihood to recommend can be determined by correlating the scores with how well a company or product is addressing those underlying customer needs. Working with companies like Pacific Gas & Electric, Southern Company, United Parcel Service and others, we have been able to show that using a rigorously derived set of attributes from a formal VOC study can dramatically improve a model’s fit; that is, it explains far more of the variation in overall satisfaction or likelihood to recommend – proof that they are a better, more realistic set of drivers than the previous set. In doing so, VOC moves customer satisfaction measurement from simply descriptive to deeply prescriptive. As one client bluntly stated at the start of a VOC engagement, “I know how to measure it. What I need to know is how to move it!”
Instead of fueling hysterical reactions over the scores, traditional VOC treats customer satisfaction as a kind of innovation problem. It provides insights to support the generation of new and better solutions to improve on the customer experience. Of course, there are no guarantees that we’ll always come up with better solutions but at least we’ll know that we're concentrating on the right problems.
This last take on VOC is well-intentioned but a bit poorly reasoned. We all know Technovangelists – those people who believe that new technology is always superior to old and that whatever you’re using now will be obsolete in a year or two! They believe that technology will solve all of humanity’s problems and that the latest thing is always the greatest thing. But reality has a way of setting in and we often find that new technologies that are great for one application aren’t always great for another. In my view, this third view falls into that camp.
The Technovangelists argue:
“Traditional VOC is obsolete; it’s been replaced by the use of social media.”
“Social media can provide all of the data needed to fully understand the voice of the customer.”
“Social media has completely changed the way that innovation is done, because it allows the customer to replace the inventor and innovate on their own.”
Can social media be used as the source from which to gather the voice of the customer? The answer so far is that it can help but with a number of significant constraints:
First, social media is a far more active phenomenon for consumer products than it is for business-to-business products. Consider dog owners. Millions of dog owners constantly think about products and services to enhance their dogs’ health and happiness and their own lives as dog owners. Thousands of them share their thoughts with anyone who cares to listen. By contrast, there are only a handful of automotive plant managers in the U.S. and they are busy people and necessarily secretive about their work, which they consider confidential and something that they prefer not to share online.7
Second, for better or for worse, social media qualifies as big data, with all of the associated benefits and challenges. Companies like social media because the data is already out there and it is virtually free. Consider the vastness of user-generated content – textual material that is there for all to see such as that found in blogs, discussion boards or social media chatter. The sheer volume demands advanced tools to decipher it. The earliest applications were at consumer products companies who wanted to follow their brand’s buzz. They developed text-mining and analysis tools to classify online comments as positive or negative and then longitudinally tracked sentiment, homing in on significant blips in the data that might require attention. For instance, if a company can get an early beat on negative chatter about a policy or issue, it may be able to correct it before it goes viral.
Many have attempted, including us, to extract customer needs from user-generated content. We’ve found that social media is quite good for problem-finding but getting to the actual need requires intelligent probing from a trained interviewer. Achieving a deep understanding of the underlying task and the difficulties in accomplishing it requires asking “Why?” a lot. Moreover, since some needs remain unspoken or unarticulated, observation is often required as well.
Undoubtedly, the tools will improve; technology always does. But for now, we’ve found two other good ways to make use of social media. First, we examine what people are chatting about in social media to develop a discussion guide of questions and topics to cover in face-to-face VOC interviews. And second, in a number of situations, we have used online communities as a source for recruiting respondents in difficult-to-reach categories. These specialized communities can be particularly useful when they are focused on low-incidence topics such as rare medical conditions, hobbies or users of an unusual product category. Even if participants’ comments on social media fall short of the level of detail needed for VOC, social media is already proving to be an interesting new source for recruiting qualified customers for more in-depth conversation. But it has not and probably will not replace traditional VOC research anytime soon.
It is thriving
Regardless of the definition or the methods you’re using to gather the data, VOC is hardly obsolete. In fact, it is thriving. Of course, skeptics will continue to offer anecdotes of failure. And even a perfectly-executed VOC project cannot guarantee a winning product or a better customer experience. But that said, there is plenty of empirical evidence of the overall success of VOC’s role in all kinds of innovation and product management.
The Product Development and Management Association Foundation conducts a long-running longitudinal study of innovation practice, the Comparative Performance Assessment Study (CPAS). In 2004’s installment, the sample of 416 companies was divided into what the PDMA called the best versus the rest. The best were the 24 percent of companies who had the best track record on innovation performance using various financial criteria such as “percent of sales” and “percent of profit from products that were less than two years old.” The rest were the remaining 76 percent of companies. The organization then went on to examine what it is that the best do differently or more of than the rest, i.e., what variables best discriminate between the two groups? In every breakdown the organization tried, VOC was always one of the top discriminators. In fact, the best were generally two-to-three times more likely to be using VOC than the rest. The most recent wave of the study (2012) confirms these findings. The best now have more than a 50 percent higher innovation success rate than the rest (82.5 percent vs. 53.8 percent) and VOC is now the single most-used market research technique, with major impact on reductions in cost and time-to-market.
In another study of over 160 companies, NPD experts Robert Cooper and Scott Edgett examined managerial evaluations of 18 different sources of insight and discovery – both their frequency of use and their ultimate effectiveness.8 Among the most effective techniques were all of the things used in voice of the customer such as customer visits, ethnography, focus groups, etc. Surprisingly, open innovation techniques were rated among the lowest in effectiveness and techniques like disruptive technologies and communities of enthusiasts were only rated in the middle – considerably less well than had been expected.
These studies clearly refute the claims that voice of the customer is ineffective or obsolete and they verify the potential limitations of social media to replace more traditional methods of data collection. When done properly, the insights obtained from VOC are often pivotal in providing the foundation for all kinds of innovation and product, service or process improvement.
A connecting point
Still, the hijacking continues. People continue to drag VOC into murky territory where it was never really intended to go. Most of this is not ill-intentioned, just misinformed. But there most certainly IS a connecting point between all of these destinations and the benefits of the journey are quite real.
Fortunately, there are still plenty of objective “passengers” on board to redirect these flights to their intended destinations. And there are plenty of good seats still available!
1 Katz, Gerry. “Hijacking the voice of the customer,” Visions, PDMA, January 2006, pages 8-9.
2 “Is VOC Killing Innovation?” iSixSigma.com. (May 4 2003). http://www.isixsigma.com/topic/is-voc-killing-innovation/ (accessed June 18, 2013).
3 While this quote has long been attributed to Ford, some have questioned whether he, in fact, ever said it.
4 Christensen, Clayton; Anthony, Scott; Berstell, Gerald; and Nitterhouse, Denise. MIT Sloan Management Review, Spring 2007: 2.
5 Griffin, Abbie and Hauser, John. “The voice of the customer,” Marketing Science, 12 (i): 1-27 (Winter).
6 Griffin, Abbie; Price, Raymond L., and Vojak, Bruce A. Serial Innovators: How Individuals Create and Deliver Breakthrough Innovations in Mature Firms. Stanford University Press, Stanford, Calif., May 2012.
7 That said, in a conversation with a group of petrochemical engineers at one of my clients, I learned that there was actually quite a bit of chatter – mostly anonymous – about how to deal with the BP Deepwater Horizon accident in the Gulf of Mexico back in 2010.
8 Cooper, Robert and Edgett, Scott. “Ideation for product innovation: what are the best methods?” Visions, PDMA, March 2008, pages 12-17.