Improving Your Customer Satisfaction System

• "We've been doing customer satisfaction research for the last five years and we have not improved one bit," said the Vice President of Sales.
  
  
• "Perhaps we should just stop doing it," offered his field liaison.
  
  
• Nothing would suit the VP better. The parent company is on his back for not improving and his franchisees are barking that the metrics aren't fair.
  
  
• "Why do we care?" The field liaison asks. "We have been measuring this forever and we have not improved one iota."
  
  
• "That would send the wrong signal to our customers and to the field that we don't care about customer satisfaction. I wish we could make it more useful," said the VP who then phones the research manager.

Sound familiar? This is the point where the market research person is told to do a "survey redesign". Stakeholders unsure or unable what to do about improving satisfaction often turn to the metrics themselves. If they could just get the right questions or scales or both, then things would really start humming.

While somewhat farcical, this conversation happens many times in many industries. Clients are really challenging themselves, and their suppliers, about the worth of customer satisfaction. Many times when things are not going the way everyone hoped, or not going anywhere at all, the gut reaction is to start changing the customer satisfaction metrics. Sometimes this is very appropriate. For example, separating diagnostics from performance appraisal functions can have very positive results. However, many times changing the metrics does not. We have to take a step back and really understand where the breakdown may be occurring.

The Chain to Success
The metric system is a small but important subcomponent of a larger system. It is embedded in a wider system of performance improvement. We call this the Performance Improvement Chain or PIC. At any point in this chain a breakdown in one of the links can occur, thus mitigating or obliterating the opportunity for channel partners to improve. Examining each link in the system can help diagnose where the breakdown may occur and what remedies might be appropriate to repair the breakage.

The perfect system would function as follows:

The first step in the system is a customer experience; in this case a channel experience. The customer goes to the outlet and buys a new plasma television or goes to a dealership for routine servicing. The customer has an experience which they're likely to remember.

We then attempt to measure this experience. The most pedestrian (and sometimes the most effective) method is for the retailer to ask, "How was everything?" Since the average large retail store owner would be hard-pressed to ask everyone this question in a meaningful way, businesses might conduct a telephone survey or mail survey or engage some other form of structured data collection. Next, the data is collected and then the information is disseminated, usually on a web reporting site.

The General Manager looks at the website, and ostensibly understands what he sees thanks to training previously provided by the home office. He notes a few customers who are upset about some "one-off" issues – some of which are legitimate and others are not. He delegates these problem resolution tasks to others for corrective action while he tackles the higher profile issues. He also identifies a systemic issue and then gathers a team to address it.

The general manager assembles the department heads in a meeting during the mid-day, mid-week lull and they examine the data together.

• The general manager looks at the very long bar in the "satisfaction with wait time" diagnostic printout of his store versus a benchmark.
  
  
• "Why are people in lines for so long?" He demands.
  
  
• "Staffing? We don't have enough people around at the right time," the HR Manager offers.
  
  
• "We don't queue people very well?" The technical specialist asks.
  
  
• "I think it is the technology…we need better scanners and registers," states the IT support guy.

Who's right? A wise person once said, "A test is worth a thousand arguments".

The team then systematically tries different interventions over a period of time. First, they increase staff during high traffic periods, and then they monitor the results using their customer experience metrics as a guide for success. They try different processes for queuing people at check-out and monitor the outcome. In short, they experiment and use the results to guide the final implementation. Finally, they implement the best solution and things improve consistently – according to their customer satisfaction metrics and ultimately their bottom line. They have completed the chain for their wait time problem and are ready to tackle the next big issue on the list.

Reflections on PIC
While extremely basic, it is amazing how many organizations do not follow this simple chain. At worst, they just adjust the measurement system hoping for a different outcome. Others get caught-up in the "law of the hammer" paradigm, which holds that when one has a good hammer everything looks like an inviting nail. This law often manifests itself in the form of re-engineering or process improvement, hiring and selection, training, compensation tinkering, or some other silver bullet. There is no silver bullet. Each problem requires a different solution or solution mix.

Let's look at the chain again. What did the general manager do? First, he had customer experiences. Next, he measured the experience. Assuming the tool was well-constructed and it measured the breadth of the experience the retailer could influence, this tool provided actionable information. The perception of the retailer efficacy in the measurement stage is an important point. Many times proprietary customer satisfaction trackers incorporate attributes that retailers either a) objectively have no control over or b) feel they have no control over. Research purists would argue we need to include every attribute influencing loyalty. I would support this position if the study's intent is to understand the drivers of loyalty and it is a joint venture between the retailer and manufacturer. Typically, customer satisfaction trackers are for the retailer to take action, not to solve global issues of product quality, advertising, or other issues seemingly out of the retailers' control. Holding a clothing retailer accountable for the quality of fabrics is an exercise in futility. After all, they don't manufacture the clothing line – they sell it.

The third link in the chain was that information was disseminated to end-users who could do something about it. Rapid problem awareness information needs to be quick in order to resolve customers' problems. Systemic issues will manifest themselves over a longer period of time and immediate turnaround is not necessary, but a depth of understanding is. In either scenario, the information must be shared in an easy-to-access and user-friendly format. Websites are great methods for disseminating information. Depending on the culture, quick and short paper reports might do the trick as well. Some make the mistake of fighting the culture in the name of progress. You will save yourself a lot of heartache by making the modality fit the need and the culture.

The next event was that the end-user understood what he was seeing. Knowledge was not innate nor was it acquired spontaneously. We in the research business take for granted the complexity we create for very busy end-users. Simplicity in design, good training and awareness, coupled with strong on-going support is important for end-users to understand what they are seeing and how to use it. The best information in the world goes to waste if the people expected to use it do not understand it.

Going back to the program design, the general manager was fortunate to have a measurement instrument that enabled him to do two things. First, it allowed him to fix customer problems one at a time. Many will stop here without looking at systematic issues. In stopping at the "fix-each-customer's-problem-stage" the larger, problem causing organizational issues may be overlooked. Admittedly, "normal accidents"1 do occur in any organizational setting – normal accidents defined as something "one-off". A better definition would be "an unfortunate confluence of unseen interactions." People make mistakes all the time and many times it stops there.

However, many times, the problem is systemic and is rooted in core institutional issues such as process, hiring, training, resource allocation, and so forth. The reason people are consistently upset is usually traceable to something the organization is doing (or not doing). In our generic example above, the general manager recognized the systemic issues from the "one-offs" and formed a team to do something about it. Having to wait all day for the cable guy is an example of a systematic issue. Detecting "normal accidents" from systemic issues is an important step in the chain in order to apply the right solution.

In our fictitious case, the General Manager had the tools and understood the difference between the "normal accident" and the systemic issue. This is rooted in good program design and just as important, good training on how to use the system.

So far, a customer had an experience, it was measured, and the end-user understood what it meant, and how to take action. If a performance improvement chain is going to fail, it is usually at this next step. That is, there needs to be some commitment to do something about it. In our anecdote above, the General Manager does commit to doing something. He also makes his direct report accountable for a) identifying the problem b) offering solutions, and c) owning the solution.

Commitment is not enough. I feel pretty committed about running every day. Others are committed to civic duties or going to church. Unfortunately, there is shrinkage between the number of people who say they are going to do something and those that do. To put it simply, "try-ers aren't do-ers". The commitment must lead to action and not just once. Many problems require repeated experimentation to resolve. Some complain, "We tried to do something and it didn't work." If problems were simple, they would not require repeated attempts for correction.

Another important undertaking at this point is to disentangle what Mark Lipsey2 calls "theory failure" from "program failure". This poses the question, was it a bad idea without the desired effect (theory failure) or was it not executed effectively or faithfully to the original "theory" (program failure)? For example, say I notice a relationship between PC failures and the presence of an I.T. field support person. I conclude that if I fire all the I.T. field support, I will resolve my PC failures. So, I fire them and see what happens. Unfortunately, PC failure increased due to my failed theory.

Now let's say I conclude that I.T. support doesn't have enough training in a particular area. I send them to training in Las Vegas and instead of attending training, they proceed to party. They return and I witness no change in PC failures (it is still high). Here I may have had the right solution (training), but it was not executed effectively (they did not attend). This framework focuses on the linkages of "did action occur?" and "did the action have the desired effect?" Drilling down on whether it is theory or program failure will allow you to modify your chain and figure out what really works and what does not.

Let's say businesses do everything I recommend above. How do they know if things improved or didn't improve? That completes the chain. We try something; it (hopefully) impacts customer experience which shows up in the metrics. We're right back to where we started.

Weak Links?
Now back to the scene at XYZ Industries headquarters. Rather than jumping to conclusions about what is broken, I would recommend doing a little meta-analysis on the overall performance improvement chain. Ask questions. The diagram below provides a break down of key questions to ask about the system as a whole. It also provides a stopping point to ask, is this linkage broken?

Each question in the chain has its own unique sets of solutions. While a topic for future articles, these solution sets are known and put into practice everyday. Before contemplating a survey redesign, I would encourage you to sit down with all your stakeholders and do some analysis on the entire performance chain. This would include channel partners (i.e., retailers, dealers), the sales division, manufacturing, the research department, and whoever has a stake in the continued improvement of channel partners. Then you can efficiently apply the right solution to the right problem, avoiding waste of valuable human and financial resources. This initial audit will allow for the solution to be applied to fixing the chain, when in fact it is broken.

• Question: How do you know if it is broken?
• Answer: If you don't realize year-over-year improvement. That's what happened at XYZ Industries.

And, that may be occurring in your organization, right now.

• ¹ Perrow, C. (1999). Normal Accidents Princeton University Press: Princeton, New Jersey.
  
• ² Lipsey, M.W. (2002). Design Sensitivity Sage Publications.

Other Related Articles from Maritz Research | In his article, "From Data to Action: Putting the Voice of the Customer to Work", D. Randall Brandt, Ph.D., VP, Customer Experience and Loyalty Research at Maritz Research, explains the value of a closed-loop process for customer satisfaction measurement and management. Brandt says, "businesses should use measures of customer satisfaction and dissatisfaction (and related VOC) to identify priority issues for improvement. Plan and implement those efforts to address those issues, and then gather new measurements to determine if actions taken are having their intended effect. Apply this process on a continuous basis to monitor and manage customer experiences in ways that will lead to desired business results."

Published by Maritz Research
Date: Volume 20 - April 2007

About Maritz Research | As one of the world’s largest marketing research firms, Maritz Research, a unit of Maritz Inc., helps many of today’s most successful companies improve performance through a deep understanding of their customers, employees and channel partners. Founded in 1973, it offers a range of strategic and tactical solutions concentrating primarily in the hospitality, automotive, financial services,telecommunications, retail, pharma workplace and technology industries. The company has achieved ISO 9001 registration, the international symbol of quality. It is a member of CASRO and official sponsor of the American Marketing Association. Based in St. Louis, Maritz Inc. provides market and customer research, communications, learning solutions, incentive initiatives, meetings and event management, rewards and recognition, travel management services, and customer loyalty programs. Maritz has a presence in 42 countries, with key offices in the United States, Canada, the United Kingdom, France, Germany, and Spain. For more information, visit .