April 2009

VoC for Doc

CIDMIconNewsletter Bob Ryan, Motorola

A Recipe for Voice of the Customer

“Management is efficiency in climbing the ladder of success; leadership determines whether the ladder is leaning against the right wall.”
-Stephen R. Covey

It isn’t possible, or desirable, to climb every wall. Customers have specific wants and needs, and it is a waste of time and money to pursue things that don’t directly impact customer satisfaction. But how do you determine which walls to lean your ladder against? You have to ask.

Since VoC—shorthand for Voice of the Customer—is often associated with Six Sigma, you might assume that VoC is really complicated and that you need to be a statistician to get it done. This isn’t the case. While it can be time consuming, the steps aren’t complicated, and a motivated person of average intelligence can achieve impressive results.

What follows is one recipe for a VoC program. It has been used on different products with different customers and has provided good results, but feel free to add a pinch of this or a dash of that to make it your own. The web site is a great resource for VoC.

Baseline Current Performance with Kano Analysis

Improvements are only improvements if they can be measured. And, the only way to measure them is to baseline your current performance. Kano analysis of your documentation set—or perhaps information set is a better term—is relatively simple and provides a clear, meaningful snapshot of customer needs and perceptions.

The first step is to sit down with your documentation set and break it down into some basic information types—no more than 10 or so. These bucket headings might include things like installation, maintenance, provisioning, alarms, commands, and so on. These are the information types you’ll be asking the customer about.

If you ask customers about books, they will tell you about books. If you ask them about information types, they will tell you about everything in that bucket. Perhaps their problems aren’t even with the documents you produce. Also, books come and go, but information types persist over time.

Once you have the information types, you need to create a short survey that asks the customer two questions about each type on a scale of one to five:

  • How satisfied are you with this information type now? (1=extremely dissatisfied, 5=extremely satisfied)
  • How important is this type of information to you? (1=unimportant, 5=extremely important)

Include just one question at the end of the survey that asks: “If we could do one thing to improve the information we provide to you, what would it be?” Customers will invariably tell you about the area that needs the most work. Don’t include this question for each information type; it just makes the survey longer.

Once the survey is complete and you have all the data, you can create a Kano chart by plotting the importance versus the satisfaction and center the axis at three (see Figure 1).

Ryan_Figure 1

Figure 1: Example of a Kano Chart

The goal is to have the customer rate each information type as having above average importance (>3) and high satisfaction (>4). Are you smiling yet? You now have a metric that is so simple that even a manager can understand it; yet, it is powerful enough to actually help you make decisions. The Kano Chart identifies the critical needs and provides a baseline score to measure the effect of future improvements.

Identify Critical Areas

In general, you are probably working on the right deliverables and your customers are pretty happy. However, the Kano may reveal a couple of areas that need some work. The information types that have high importance and low satisfaction are the areas where your ladder may be leaning against the wrong wall.

Select the 20 percent of your information types that most obviously need improvement. Sit down with your team and talk about the key users of each type of information. If you gathered some demographic information from your original survey, you might already have the data you need. For example, if installation information is a problem, you’ll need to talk to customers who perform installation procedures.

Armed with the Kano and your demographic assessment, arrange customer interviews, face-to-face if possible. Such meetings are, of course, easier said than done, but direct interviews are much more effective than surveys. Also, the Kano chart and some well selected verbatim comments will open a lot of doors. Don’t underestimate the power of displeased customers.

Interview the Customer

The interview is the most critical part of a successful VoC activity. Frankly, your customers have better things to do. You need to show them that you value their time by being prepared and getting what you need on the first try.

Start by creating an interview guide. Use what you have learned from the Kano analysis to narrow down a few high-level themes, for example: ease of use, completeness, timeliness, accessibility. Then come up with a few probing questions around each theme. Here are some guidelines for creating good questions:

  • Don’t ask leading questions—The goal is to understand what the customer wants, not tell them what they want.
  • Ask open-ended questions—The more the customer talks, the more likely you are to discover the source of their dissatisfaction.
  • Why? Why? Why?—Probe for greater detail when you get a vague or incomplete answer. Keep asking ‘why’ until you are satisfied.
  • Practice—Get one or more people to play the role of the customer and interview them. Practice will polish your interviewing skills and reveal potential problems with your questions. Work the kinks out of the interview internally before you sit down with the customer.

When the time comes to conduct the interview, do not try to both ask the questions and capture the customer’s feedback. It is not possible to write and listen at the same time. Have one person conduct the interview and one person capture everything the customer says—a spreadsheet works great for this. A digital recorder will also work if having a scribe present is not an option.

Perform a KJ Analysis

Take a moment to pat yourself on the back. You have met with the customer and now you have data—lots and lots of data. Don’t panic. KJ Analysis, named after Japanese anthropologist Jiro Kawakita, is a time-tested method for distilling a large number of seemingly unrelated comments into a clear set of critical customer requirements. The goal of performing a KJ Analysis is to ensure that your organization is targeting requirements that are New, Unique, or most Difficult (NUDs). Focusing on the NUDs will provide the greatest customer-satisfaction bang for your buck.

Schedule the KJ Analysis meeting to occur as soon after the interview as possible. At a minimum you need your team, a conference room, and a stack of sticky notes. If your team is geographically dispersed, you might want to consider using mind mapping software and an online meeting. The example in Figure 2 was produced with Mind Manager from Mindjet Software.

Ryan_Figure 2

Figure 2: An example of KJ Analysis

Here is the basic flow of a KJ Analysis meeting:

1. Write all customer comments (voices) on sticky notes (one per note) and stick them on the conference room wall.

2. Have your team silently read each note and arrange them into groups of similar comments.

3. Create a requirement statement for each comment group.

4. Focus on the NUDs by setting aside any requirements that are old, common, or easy.

5. Have the group vote on the remaining requirements to rank them in order of impact.

The silence in step two is important. If you allow people to speak, there is a pretty good chance that the process will end up being driven by the strongest personalities. It is OK if sticky notes get moved many times; the most important thing is that the process is a collaboration.

Don’t discard the old, common, and easy requirements. Many of these items can be addressed as time permits in future documentation revisions.

If you end up with more requirements than your group can handle, you’ll need to rank them. Provide each team member with a limited number of stickers (those garage sale price tags work well) that they can attach to the requirements they think are most impactful. Make sure people have fewer votes (stickers) than there are requirements to vote for.

Verify Findings

The only VoC step left is to go back to the customer and present your findings and proposed improvements. Their feedback ensures that you have understood the underlying issues and have defined solutions that the customer really wants. Be prepared to make some adjustments. Also, make sure your organization is committed to providing the necessary resources to complete the requirements before you present them to the customer. Don’t promise anything that you can’t deliver.

Lather, Rinse, Repeat

The corporate term is Continuous Improvement, but I think the old shampoo instructions sum it up nicely. Repeating the Kano and KJ Analysis each year will provide your group with a clear performance metric and will help define organizational goals. Another side benefit is that it will shield you from unfounded attacks and provide you with the ammunition you need to get the tools and resources you need to do your job. Most importantly, it will improve customer satisfaction.

About the Author

Bob Ryan_bw

Bob Ryan

Bob Ryan is currently a Technical Publications Manager at Motorola with nearly 15 years of experience in technical communications.  He is a certified Six Sigma Green Belt and has completed Voice of the Customer (VoC) projects within the documentation group with a certified cost savings of greater than $100,000 to the corporation. Though he became a manager in 2000, he has never fully relinquished his role as a writer and continues to author online software documentation. Bob has a B.A. in Journalism from MacMurray College.