There’s often a lot of discussion and debate on what to measure when assessing IT service desk and IT service management (ITSM) performance. The ITIL books offer a wealth of possible metrics, key performance indicators (KPIs), and critical success factors (CSFs) – so many that people shouldn’t be using all of them. As with ITIL as a whole, use what works for your organization. And remember that sometimes less is more, with a commonly adopted measurement being customer satisfaction.
Customer satisfaction is definitely a useful metric but, in my opinion, it seems to be difficult for organizations to measure and understand. With it only useful if used in the right way – for instance, asking the right people the right questions, and ensuring that the responses received are truly representative.
This blog serves as a quick guide to help you create reliable measures for assessing customer satisfaction, with the following six tips:
1. Select and Get the Right Sample
It starts with getting people to answer your questions. If the total number of customers is small, you can ask everybody. Otherwise, make sure that you have a random sample selected. Never ask only the people who have contacted the IT service desk, it’s a highly misleading and select group of people. It doesn't represent your entire customer base as it’s only the ones that want, or are willing, to call the IT service desk and it’s heavily biased toward those who contact the service desk often.
2. Incentivize Feedback
The best way to get everyone to answer your customer satisfaction survey is bribery. Give people something for their trouble, it usually works.
When I was running conferences, we had a simple system – when people left the conference, they got a gift for completing the customer satisfaction form. Our response rate was usually a bit over 100% as everyone wanted the gift even if they were not strictly eligible for it. For example, one of the gifts was a decent umbrella, and I still have mine in the car.
Use your imagination, something as small as a free soda or chocolate bar is a strong motivator if the survey is easy to answer.
3. Ask the Right Questions
This is a key success factor: the questions. Do not ask too many questions.
I'll repeat that: DO NOT ASK TOO MANY QUESTIONS.
And do not ask too-complicated questions. Sadly, consultants love to invent very complicated questions just to show how clever they are. In my experience, open questions are the most valuable ones. Replies to open questions can be very valuable, especially if many people are independently saying basically the same thing. Analyzing open questions is a bit more time consuming than just calculating percentages but it’s worth it.
Here is a simple model with three questions, one closed and two open:
- How did you like our service, please use a scale from 10 to 1 where 10 is excellent, and 1 is unacceptable: __
- What did you like most about it?
- What could have been improved?
In the conference example, we had the dilemma that there were often more than thirty speakers and the form would have become complicated if we had asked about every speaker. We solved this by asking the respondent to name the three best speakers. The questions worked quite well. We could pick out the speakers we would like to ask to speak at our events again. Generally, almost all speakers got mentions in the top three list question.
4. Understand That Satisfaction Does Not Equal Value
Value is important, the whole idea behind any service is that it creates value to the customer(s). Satisfaction is good, but it’s not the goal.
5. Avoid Simple Percentages
Do not use simple percentages such as satisfied responses or net promoter score (NPS). Percentages lose data and NPS is very unreliable and unsuitable to many services.
But simplicity is important. Don’t use overcomplicated methods like SERVQUAL, where you must assume that all respondents understand the questions in the same way. Keep it simple.
6 Understand the Cost of Bad Surveys
You might think that these methods are expensive. But they’re not – what’s much more expensive is running a bad survey and getting misleading results.
So, that’s my six tips for better customer satisfaction measurement for the IT service desk. What would you add?