The lack of a robust voice of the customer (VOC) process, no proper training in applying VOC in project selction and evaluation, and VOC systems that focus on voice of the business issues keep many Six Sigma companies from being truly customer-driven.

By John Goodman and Bruce Hayes

Many Six Sigma efforts are not really customer-driven. Too many companies pay only lip service to the needs and perspectives of stakeholders and customers when selecting and deploying Six Sigma projects. There are three main reasons for this:

• First, Six Sigma-trained personnel, and many times even Black Belts, are not properly or specifically educated about how to use the voice of the customer (VOC) to select and evaluate projects.
• Second, most companies lack a robust VOC process, resulting in little effective, data-based input.
• Third, many of the VOC systems that do exist are really disguised voice of business (VOB) systems, which focus primarily on the cost of customer problems and other VOB issues, i.e., efficiency, productivity, internal quality.

Lack of VOC Training or Strong VOC Processes

Conventional Six Sigma training may spend only a few hours on customer survey design and analysis while spending multiple days on project costing, defect tracking, return on investment (ROI) calculation, internal VOB and statistical tools. Tools for customer data collection, prioritization and distillation are rarely studied. This imbalance leads to what often are referred to as “requirements failures.” A review of one corporate Six Sigma training program by TARP, a leading customer service research firm, revealed a disturbing fact: The program, which had been presented to every executive and manager in the billion-dollar company, failed to mention the customer at all until page 37 of the training manual, and then only one other time, on page 127, where customer requirements were assumed to be known.

Many companies assume that they have clear, valid data on the requirements of their customers. But TARP’s benchmarking indicates that VOC processes in 70 percent of companies are ineffective. A Standish Group study shows that 50 percent of the failures in technology projects are requirements failures. This can be costly. Poorly designed VOC data collection systems and the lack of VOC tools are estimated by the National Institute of Standards and Technology (NIST) to cost U.S. corporations nearly $100 billion a year in failed technology projects. According to NIST, 31.5 percent of software application projects, on which companies expend $300 billion annually, are flat-out canceled.

In addition, the VOC input that is used generally does not take into account customer loyalty, and its impact on revenue. This lack of customer input results in three problems – the wrong Six Sigma projects are selected, Six Sigma is used for cost cutting (which hurts customer loyalty) and projects tend to have a limited focus for process improvement. Authors Brian Swayne and Brent Harder reported in their article “Where Has All the Magic Gone?” (Six Sigma Forum Magazine, May 2003) that cost savings was the justification for almost all projects examined in 11 companies.

Heavier Focus Is Put on VOB

Six Sigma improvement projects use the Define, Measure, Analyze, Improve, Control (DMAIC) roadmap, which drives bottom line results (cost savings) in most cases. Although balance between VOB financial performance parameters and VOC is taught to Six Sigma practitioners, a heavier focus on VOB is often the norm. Six Sigma product, process and technology design (or re-engineering) projects use the Design for Six Sigma (DFSS) roadmap and are more likely to drive top line revenue growth opportunity, market share improvement and/or innovation in design.

Customer balance through data is needed in both. An effective VOC process will identify and distill, not only the problem areas, their cost and their revenue implications, but also the causes of the problems. And many of those causes are not production defects, but often are either marketing and sales defects which incorrectly set customer expectations, or customer-caused defects, e.g., mistakes. Both types of problems can be prevented by better communication with the customer rather than by re-engineering the production process. Customer education has been found to improve the quality of the customer’s input to the service delivery process by as much as 50 percent (in terms of reduced problems) with a 25 percent increase in loyalty. The downside of failing to use an effective VOC data collection process is that ROI is underestimated, and damage can and will be done by improving efficiency but decreasing customer satisfaction in the process.

Requirements for a Useful VOC

The following eight factors are required for an effective and institutionalized VOC process.

1. There must be a clear owner of the VOC process who is responsible for managing the process, getting and controlling input from all touch points, and assuring that issues raised are clearly assigned to an individual executive within the organization. Further, multiple “listening posts” such as focus groups and interviews must be conducted at multiple levels to assure the elicitation of valid useful data. Once this data is collected, it can be used in two ways – at the macro level to drive Six Sigma project selection and immediate improvement opportunities, and at the micro level to provide specific requirements for individual DFSS projects.

2. There must be a unified feedback and quality data collection plan for all touch points and internal sources of data describing the customer experience. The methods and classification schemes used in all systems must be compatible to support integration. For data sources which are not representative of the marketplace, such as complaint systems, there must be an analytical means of extrapolating the information to the marketplace as a whole. For example, answer the question: “For each complaint we are getting about this type of problem, how many other customers in the marketplace are encountering the same one but either are not complaining or are going to another touch point?” In many cases, the answer is 10 customers for each complaint or even 50 customers per complaint received.

A further part of the VOC data collection activity should be mechanisms to gain an understanding of the current and evolving motivations and expectations of customers. Focus groups and interviews with early adopters to identify future needs and channels of communication will allow prediction of problems and needs before they turn into problems. Launching appropriate projects to address the long-term needs of the customer and the business are valuable. All of these points will significantly enhance a Six Sigma practitioner’s ability to accurately select and charter important customer-facing projects and assure the availability of statistically valid data.

3. A unified, single picture of the customer experience needs to be provided based on data gathered from multiple sources and which integrates all phases of the customer experience. If the data is not integrated, management and Six Sigma teams are provided with multiple pictures of the experience, resulting in conflicting opinions on where the biggest problems are. Also, multiple data sources saying the same thing help eliminate the politics associated with delivering bad news. As most Black Belts know, multiple data sources, once validated for accuracy, improve the probability and speed with which solution sets can be developed and implemented.

4. The VOC reporting process must be easily accessible, visible to all managers and actionable. Actionability is produced when the data flags a specific defect or unfulfilled need along with enough information to determine its root cause. This root cause analysis should examine customer expectations to see if they are realistic. Educating the customer is a way to improve the customer’s contribution to “production” of the product or service. As mentioned, requirements failures are the manifestation of a poor VOC process. Complaining customers and other customer dialog tend to drift into the solution space as opposed to the actual problem space. Trained DFSS Belts are skilled at using data to elicit further conversation, mining both VOC quantitative data and “comment data” for context, meaning and root causes.

5. The VOC process must identify the complete cost and revenue implications of the problem or opportunity. The ranking of issues by a comprehensive estimate of cost and revenue impact helps to establish priorities in selecting Six Sigma projects. Keep in mind that there is significant revenue damage even if the customer encounters the problem but does not ask the company to do anything about it (so that there is no immediate internal cost or rework).

The revenue estimate must be conservative enough that the CFO will accept it. A critical part of this calculation is to use a conservative estimate of the revenue value of the customer. TARP recommends using the revenue flow of no more than two or three years at most. No CFO will accept a lifetime value based on 10 years of purchases. The market is too dynamic for anyone to be certain if the market will even exist 10 years from now. An additional best practice is to verify predictions of customer attrition and problems by checking actual customer behavior 6 or 12 months later to show that customers usually do what they say they will. This type of data is valuable to help Six Sigma Belts calculate the ROI of their projects in areas which are traditionally documented as “soft savings.” Having this “hard data” and the CFO buy-in will accelerate the top line revenue opportunity attributed to the Six Sigma program, increasing its overall effectiveness significantly.

Further, the revenue impact should also take into account the impact of word of mouth, which in many companies contributes between 20 and 70 percent of all new customers. Service and quality are major contributors to the word-of-mouth management process. Finally, the analysis should consider the cost of responding to the incorrectly set customer expectations that may take place in a completely different part of the organization.

6. The VOC process must identify what level of performance would be acceptable to the customer, therefore suggesting a goal that can be assigned to a company dashboard. Once the goal and performance tracking are in place, the effect of specific actions can be monitored and optimized, accountability and responsibility established and improvement actions controlled.

7. The VOC process must track and report progress in a highly visible manner. One of the most prevalent weaknesses TARP has seen in VOC processes is that they fail to clearly ask, “Did the problem get fixed?” That which gets measured and reported gets attention. This fundamental feedback loop must be institutionalized in the measurement process to assure full closure of issues reported and requirements stated.

8. The VOC process must be linked to organizational incentives to reward those who take action, get the problem fixed and enhance revenue while reducing cost. Normally a robust Six Sigma program has such incentives in place tuned to the culture of the organization.

Assuring a Robust VOC Process Is Available

In companies where these eight factors are not in place (which is most companies), the Six Sigma team must take three actions to assure that a robust VOC process is available to support its efforts.

First, there must be an effective program of education on the principles of customer service in order to obtain the buy-in of finance, product and business management. Also, there needs to be a common understanding of the likely behavior of customers when they encounter a problem, and how to measure and put a value on that behavior.

Second, the VOC process must be developed enough to provide information about the customer experience across the entire life cycle to assure that the most important problems, issues and requirements are highlighted and acted upon. Ideally, the information should come from multiple sources – some solicited from customers, some from employees (like sales representatives or delivery staff or accounts receivable staff), and some from unambiguous internal data systems, like returns and credits. Multiple sources allow for the biases inherent in all quality data sources. Multiple listening posts, user groups, surveys, call center data and other techniques should all be used in a collaborative way to bring strength and correlation to the data and subsequent conclusions.

Last, the finance staff must be educated on the methods of conservatively measuring loyalty and word of mouth and converting them into revenue implications. Also, there must be agreement on a conservative estimate of the revenue value of a customer.

Using the VOC to Select Projects

The following steps should be followed in selecting DMAIC projects (short-term improvement):

1. Identify candidate projects based on areas of significant dissatisfaction and/or complaints.
2. Identify at least three data sources, one internal and two external, which describe the the frequency of issues or problems in the marketplace. If such data does not exist, estimate the frequency based on an extrapolation of the number reported at particular touch points.
3. From those issues, select the ones that seem to be doing the most damage by roughly estimating the revenue damage (based on impact on loyalty and word-of-mouth referrals when the problem occurs [severity], frequency of occurrence, and value of customer).
4. Identify the internal cost of handling those issues and the revenue impact, both short- and long-term.
5. Decide whether the issue can be fully eliminated by a process change or mitigated after occurrence, and the relative cost of each action.
6. Decide which issues provide the best cost/benefit, and select the projects for action.

The following steps should be followed in selecting DFSS projects (long-term product/process design):

1. To identify or validate design ideas, pinpoint trends in what customers are saying about their product or service. Give special attention to compliments or stories of customer delight because these indicate where the customer has received extra value, often at little extra cost.
2. As in DMAIC, look at the damage done via specific issues customers encounter and ask what long-term changes in process or product would eliminate those unpleasant surprises or decreased value.
3. Identify the scope, cost and long-term VOC/growth impact of the ideas.
4. Determine which candidate projects are consistent with the desired brand promise.
5. Select those projects which fit the brand promise and also have high cost benefit.

Estimating the Payoff

Use the same market impacts to estimate the payoff of the project after implementation. Identify the reduction in problem occurrence or increase in delight and perceived value due to the Six Sigma project, estimating the customer problems avoided and the increase in loyalty and positive word-of-mouth referral level.

Also, estimate the reduced service workload due to reduced problem occurrence. Archive the project results, communicate the success and drive a shared understanding of what, how and why the project was successful across the organization.

Conclusion: Demand an Actionable VOC Process

Six Sigma Champions and Black Belts often make project selections with little VOC input because they find that available VOC data is fragmented, contradictory and fuzzy on revenue impact. Rather than ignoring VOC and falling back on selecting projects which address only cost reductions, Six Sigma professionals should demand that departments which collect VOC such as contact centers and market research, create an actionable and coherent VOC process. Perhaps the development of an actionable VOC process would even make a good Six Sigma project.

About the Authors: John Goodman is president and a founding member of TARP which was started in 1971. TARP’s research instigated the use of toll-free phone numbers for customer service by companies like GE, American Express and Toyota. Mr. Goodman specializes in developing voice of the customer processes that receive support from the CFO. He can be reached at [email protected]. Bruce Hayes is co-founder of Six Sigma Advantage, where he focuses on business development, executive coaching and training, and curriculum development in Six Sigma for technology. He is an operations professional and consultant with more than 29 years of experience. As a senior executive at Motorola, he was one of the key contributors to developing and “operationalizing” Six Sigma. Mr. Hayes also was co-founder and president of a leading conventional Six Sigma consulting firm. He is now engaged in porting his experiences into the technology and software arenas. Mr. Hayes can be reached at [email protected].