If you’ve been following this Blog, you know that we have been discussing variability: how the products and services we provide can be better or worse at any given moment. We then created a tool for identifying and quantifying the variability in our services, the P Chart (take a look at the August 27^th Blog, “4th Sigma: P Charts Made Simple”). That’s a pretty good start! Now, that you have a way of finding the variability in your services, what do you do next? Well, next you need to find out how variability affects your clients. And that, is what today’s blog is all about!

To start, let’s go back to another old discussion: client complaints. You probably have some process to deal with client complaints. When a client calls or emails with a service issue, you probably record that instance. You may reply with a feedback form or a survey, or you might have someone talk to the client and find out more about the issue. In any given month how many of these complaints do you see? Perhaps 5, or 10, or 100… maybe zero for some months? Did you once have a lot of complaints, and now they are declining? Does that mean that your service is improving? Unfortunately, if you just look at complaints you can’t tell if your service is improving. While it does not seem logical, a decrease in complaints may mean that clients are becoming increasingly negative about your service. However, if your metrics are improving and complaints are dropping, why would clients be turning negative?

Think about what complaints tell us. Details of a complaint provide the clients perspective on a problem with the service. But what do we learn from the complaint itself? We learn that the client, even after they have a problem, still believes in you and in your service. If the client doesn’t believe that you are listening, or that complaints don’t lead to improvements, or that complaints lead to retribution, it wouldn’t make any sense to enter a complaint… would it? However, there is a way to find out the meaning of your client’s silence. Go back to recent P-Charts. Let’s look at on-time-delivery as our example.

In previous blogs, I’ve said that on-time-delivery (OTD) is the most frequently mis-measured metric. We usually make all sorts of adjustments to OTD before we report it (it was just late by a few minutes, we re-negotiated the deadline and no one complained, if we counted OTD at shift change our metrics would be too low, etc.). Assuming that your metrics are accurate and you are reporting REAL OTD, then count the total number of times you missed the deadline. Got it, OK? Now, look at the total number of complaints you received. How do they compare? Remember, this is just one metric. Now add up all the missed metrics (quality errors, missed deadlines, and any other SLA’s), what is the total count every month?

Every firm, and every department is different when it comes to complaints. Furthermore, complaints can be recorded differently. For example, if you ordered 10 new computers and 5 were installed incorrectly, how many complaints would you get? If they were installed individually over a few days you might see a maximum of five complaints. If they were all installed on the same day and someone on the client side coordinated the installation, there might be one complaint for the installation (with the details of the complaint indicating that there were five problems). Alternatively, if there was a coordinator but computers were installed on multiple days, then the coordinator might enter two or three official complaints, but might stop after that… especially if the type of complaint (wrong version of Microsoft Office?) was the same for all instances. So, the way that people think and the way that you collect this information will dramatically alter your complaint levels. Having said that, let’s take a look at the ratio of recorded complaints vs. missed metrics:

• 75% – 100%: If every mistake and missed metric leads to a complaint, they you have very good communication with your clients! However, it is unlikely that you will get a response level this high. A good client will forgive small errors that happen only occasionally, and may even be unaware of minor service level issues. For example, if a copy center agreed to drop off work on a client’s desk by 1:00, but the client doesn’t get back from lunch until 2:00, the client may not know nor care that it was dropped off at 1:30. So, kudo’s if complaints and  missed metrics match this well, but I wouldn’t expect it.
• 25% – 75%: This is more likely to be the right ratio. You need to test this assumption based on how you collect data, but it’s a good starting point. Regardless of what you find is “normal” your next step will be to increase the level of feedback until you find the natural “ceiling” for complaints.
• 5% – 20%: In this range, something is wrong. Perhaps your clients don’t understand that complaints are a good thing, and something that you need in order to operate your service. Alternatively, there may be real barriers to reporting complaints. Your P Charts will show you the degree of the discrepancy between the number of complaints you should expect, which you can compare to the official complain level. You then need to go back to the data to find the specific clients who should have complained, but didn’t. Find clients who have frequent reasons to complain but don’t (work repeatedly missed SLA’s, but never entered complaints). Understanding this silent population will tell you a lot about how your services are perceived.
• Under 5%: Receiving only one complaint for every 20 (or 50 or 100, or more) reasons to complain, heavily skews the data. In this range, client complaints may appear to be random and provide little meaningful information about your service. Such low feedback levels are dominated by a few fearless (or cranky) individuals who speak up. By definition these individuals are different than their peers; not surprisingly, their feedback usually does not reflect your service as a whole. You need to get more feedback. At these reporting levels, it is not unusual to find that the clients either have a very negative view of the service or they believe that individuals making complaints will be punished (usually that they will get slower response times, still lower levels of service, etc.).

We now know that a lack of official complaints could mean that your clients love you, or it could mean that they don’t trust you or believe that you will improve. If you want to find out the truth the answer lies in variability, and P Charts can show you what you need to find out. You’ll need to go back to the original data sources to find specific clients to talk to, and some may be very reluctant to provide information. Not everyone needs to participate, and at first you may not have the manpower to expand your customer service to handle this new definition of a “client with a complaint.” Identify the resources you can dedicate, and then target the segment of your client population that will yield the best feedback.

It is always very difficult to go after clients for negative feedback. The process makes many clients uncomfortable, and what they say can be very hurtful to you and to your staff. It often feels easier to leave these issues alone. But, when they’re left alone they always grow and become uglier. Get out there and start filling up that silence with truthful feedback… and that’s my Niccolls worth for today!