Competing for Performance

Motivation is one of the great challenges of help desk management. Keeping good morale among analysts and encouraging them to strive for continuous improvement is difficult in a line of work where there are so many customer requests, and most of them come from frustrated end users who have already tried everything to resolve their problem on their own.

One great enabler of positive morale in this environment is to set individual goals for analysts and create a competitive environment. By seeing each other's metrics, each analyst will see objective measurements that compare them to their peers. For most, this is enough motivation to strive for continuous improvement. For those who are not driven by these types of goals, you as the manager will still be able to see where they stand in relationship to their peers and coach them toward improvement.

Many managers are squeamish at first when considering competition. How will it affect the people on your help desk? Will it destroy teamwork and be counterproductive?

There are several ways to make a smooth transition into a competitive, metrics-based culture. One approach is to create a comparative report but replace all of your analysts' names with a label, such as Analyst A, Analyst B, etc. Each individual only knows what their label is, and can see where they stand in the ranking.

Another approach that has been used successfully is to generate both individual metrics and team average metrics. Reporting these on the same score card and distributing to each analyst only his or her own metrics, compared to the team averages, lets each analyst know if they are above or below average. As everyone strives to be above average, the average rises higher, encouraging more competition and continuous improvement.

When each analyst sees only his or her individual scores against the team average, you can also set up a second competitive initiative -- Star of the Month. By making a big deal about your top performer and sharing this individual's metrics, your top talent will strive to become the star. This higher band of competition will take people with a lot of drive and initiative to the next level and reap tremendous rewards. Performers who are below average may not even consider reaching this level, but they will still have the intermediate goal of the team average to move them up to above average.

What metrics should be included in the competitive mix? Typical measurements include: average speed of answer; first call resolution; percent of schedule adherence/availability; case backlog count and follow up; and customer satisfaction. Focus only on the most important metrics so that analysts keep a clear focus and are able to perform.

Let the competition begin.

The Help Desk: Dumping Ground or Focused Performance?

Many help desks become the dumping ground for security, administrative tasks, training, non-support communication, change management, and ID setup. In addition, many second or third level support teams use the help desk as a shield to keep away from customer interaction. Most of these things are distractions that cloud your main purpose and prevent you from serving your customers effectively.

The distracted help desk causes stress for the managers and team leads that oversee it and requires a highly reactive management style. The successful help desk, on the other hand, focuses on the essential service desk disciplines and resists the urge to heap unrelated tasks on an already lean team of support professionals.

Staffing a help desk is seldom senior management's priority, so discussions are often more about outsourcing rather than adding the needed head count. This increases the importance of maintaining or reducing costs while improving service levels. To do this requires focus.

To lead your support team successfully, focus on these fundamentals to raise help desk performance to optimum levels.

1. Real-time handling of support requests
Spend time each week looking at call volume and email workload on an hourly basis. You will find trends for each day of the week and each hour of the day that help guide your scheduling. Analysts must clearly understand your expectations on phone availability time, email time, follow up time, and any special projects you assign. By using a schedule that you create based on call and email volume, you will make dramatic improvements in customer satisfaction. To be precise about your scheduling, use Erlang C calculations to predict the number of analysts needed per hour.

2. First call resolution
Few things will delight your customers more than resolving their support request when they call, or within a short time of their email. Your efforts to improve first call resolution will help them get on with their day and improve your interactions with them. In addition, it will lighten the load of the support teams that receive escalated cases. To make the most of your limited training time and money, focus training on improving first call resolution on an analyst-by-analyst basis.

Your measurements should tell you which types of cases you receive most. If your categories are too broad, you may need to look a subcategory to make these measurements meaningful. Working in order from the most frequent types of cases, tailor a training plan for each help desk analyst that will improve his or her resolution in the case types that make up 80% of your cases. Supplement this with a well-maintained knowledge management tool to further improve first call resolution (see Creating a Support Wiki for one approach).

3. Open case follow-up
Work diligently on closing open cases that could not be resolved on the first call. Set expectations that each analyst should update each case with status information on a periodic basis. Unless you can measure this and show analysts how they are doing, many cases will be left without updates and will frustrate your customers. Set service level expectations with your level three support teams and hold them accountable as well.

4. Process improvement
Spend time each week identifying, planning, and implementing process improvement opportunities. Research new technologies, training approaches, and trends in running a help desk. Automate all of your metrics. Innovate. This will allow you to improve the quality of your services while decreasing costs.

5. Career development
While you are focusing on performance and the quality of your services, make sure that you are satisfying another critically important customer -- your help desk analysts. They will provide the best service when they enjoy the work, are treated fairly, and are given growth opportunities. If you staff your own help desk, you need to think of yourself as a talent machine. You take entry-level people in, train them and help them to develop career growth potential, and after a few years, you help them find work in another department of your company as a result of the skills they learned on your help desk.

In summary, focus on the fundamentals. Say no to work that steals momentum from these essential areas, and create an engaging, professional support services team.

-Steve McElwee

Case-by-Case Customer Satisfaction

One of the most revealing ways to measure your help desk or service desk performance is to survey your customers. There are many approaches to surveying, but some are more effective than others.

Many companies will annually send a customer satisfaction survey to the whole company to gauge performance. There are several problems with this approach. First not everyone surveyed will have used your help desk's services. This will either make your percentage of responses low or add irrelevant responses to your data. Since you only request this information once in a while, you will be tempted to cram too much information into the survey. Often these surveys lack focus and do not give you a clear picture of performance.

A better approach to help desk customer satisfaction measurement is case-by-case surveying. Each time a case is closed, meaning the problem was resolved and the resolution was already communicated to the requester, a short survey on the case is sent to the requester. Some companies limit the number of surveys sent to the same end user in a given month.

Here is a sample survey, based on critical to quality (CTQ) goals that you may have already established.

Please describe your satisfaction in the following areas for your recent help desk case:

Availability - how quickly we responded to your call or email:

1. Very satisfied
2. Somewhat satisfied
3. Somewhat dissatisfied
4. Very dissatisfied

Turn Around - how quickly your request was completed:

1. Very satisfied
2. Somewhat satisfied
3. Somewhat dissatisfied
4. Very dissatisfied

Courtesy - how well we listened and empathized with you:

1. Very satisfied
2. Somewhat satisfied
3. Somewhat dissatisfied
4. Very dissatisfied

Communication - how well we communicated with you on this case:

1. Very satisfied
2. Somewhat satisfied
3. Somewhat dissatisfied
4. Very dissatisfied

Comments

When reporting on your results, you can measure customer satisfaction for each of your CTQ goals. Simply report on the percentage of responses that were either 1 or 2. For example:

Customer Satisfaction % = [Number of 1 and 2 responses]/[Total responses for CTQ]

Taking this to the next level, you can review the comments provided by your customers. By reporting on comments where one of your CTQ goals was 3 or 4, showing dissatisfaction, you will find suggestions for improvement. Reporting on your 1 and 2 responses, showing satisfaction, will show you what you are already doing well. This is great information for continuous improvement.

Another way you can use the results is to tie the case ID back to the analyst who took the call. This is a good way to balance individual analyst performance with other internal metrics, like first call resolution. For this it is important to embed the case ID in the request for the survey, since end-users may not remember their case ID.

If you want to be successful in customer satisfaction measurement, you need to automate everything. If surveys require time from other important work, it will quickly go by the wayside. Automate both your requests for completing the survey and your collection and reporting of the data. Some help desk management software will do this for you. If yours does not, consider customization that will enable it -- it's worth the investment.

Good luck with your surveying. Hopefully you will be pleasantly surprised by the results.

-Steve McElwee

Erlang What?

So what's all the hype about Erlang C? What does this have to do with leading a help desk?

Erlang is a unit of traffic measurement. It is measured by looking at the usage time for all resources divided by the total time interval. It is used for telecommunications design, architecture, and call centers. It's a great fit for the help desk as well.

On the help desk, Erlang C is a way of calculating the number of analysts needed for each hour of the day. Inputs into the formula are simply:

• the number of calls received in the hour you're looking at
• the average duration of these calls (in seconds)
• the acceptable average delay (in seconds)

Although you can calculate the number of analysts for each hour on your own, you can simply use one of the freely available Erlang C calculators available on the web, like http://www.erlang.com/calculator/erlc/.

Remember that for each hour of coverage, you'll need to do a separate calculation, since call volumes vary throughout the day. You can use this information to schedule analysts. Depending on the variability of the volume of calls, you may need to consider using part-time analysts during certain hours.

Why Knowledge Management?

One of the worst labels that is often stuck on help desks at many companies is, "helpless desk." This name is a reaction to what end users feel when they call. Too often they get an analyst who knows less than they themselves know about supporting the product. As a result, the analyst takes up some of the end user's valuable time, only to create a ticket and assign it to a more specialized support team.

First call resolution and resolution within the help desk are two measurements that will show you how you're doing. If your first call resolution is as low as 50%, you'll find frustrated users, since one in two don't get the help they need as quickly as they should. Some of the best help desks have first call resolution that exceeds 85%.

Since many help desks have high employee turn over, the learning curve to prepare an analyst for optimal performance is costly and requires a lot of time. This also reinforces the "helpless" stigma.

If you find yourself on the low side of first call resolution, you may need an efficient knowledge database for creating, auditing, and using standardized solutions to end user problems. This will not only help boost first call resolution, but will also help reduce the detrimental effects of turn over.

There are several keys to successfully implementing a knowledge base for boosting first call resolution:

• Make sure it's very easy for your analysts to use. If it takes more time accessing the database than trying to figure out the problem on their own, your analysts won't use it. Having a search engine and browseable categories can help with this tremendously.
• Define a process for expanding the knowledge base. Defining roles for who authors, who approves, who edits, who audits, and who uses the database is crucial. Making these roles part of annual objectives can go a long way to ensure success.
• Create an annual review process. Have an auditor regularly look at solution articles that are used infrequently as well as commonly used articles. Create data fields that track the audit date of the articles so you can ensure that your analysts are using good information.
• Gain support from support teams to which tickets are escalated. Support teams don't like repeatedly solving the same problem. You can help them to win by having them look at repeated requests and asking them to contribute solution articles for the knowledge base. The more you can get involved, the quicker you'll see results.
• Assign someone the role of editor to ensure consistency and to create and maintain the classification structure.

There are both commercial and open source knowledge bases available to help you get started. Here are a few open source (free) projects:

• Firefly
• Memento
• CompStrm
• Exteca

Get started now, and watch your end users' perspectives change for the better.