Set Service Level with Call Center Cost -- and the Customer -- In Mind
Empowering contact center excellence for 30 years!

Set Service Level with Call Center Cost -- and the Customer -- In Mind

The number of staff you need to handle contacts and the schedules you produce should flow from your service level objective. Imagine that, in a half-hour period, you’re going to receive 50 calls that last an average of three minutes. If you have only two people to answer the calls, the delay time for most callers will be long, and you’ll probably have high abandonment. As you add people, delay times will drop.
How many people should you add? Enough to reduce the queue to an acceptable level for you and your callers. In other words, the answer to that question becomes your service level target, and you won’t be able to achieve your target without the correct level of resources.
There is generally no “industry standard” service level that you can hang your hat on. (There are some exceptions — e.g., service levels for utilities may be regulated.) The optimum service level is affected by a host of factors, including the value of the call, fully loaded labor costs, trunk costs and caller tolerances. An industry standard would have to be based on all call centers having the same values for these things.

Why There Is No “Industry Standard” Service Level

The optimum service level is affected by: 
• The value of a call 
• Labor costs 
• Telecommunications costs
• The seven factors of caller tolerance 
• The organization’s desire to differentiate products or services by the level of service provided
The Correct Service Level for You Is the One That:
• Meets customers’ needs and expectations 
• Keeps abandonment at an acceptable level 
• Minimizes agent burnout and errors 
• Minimizes expenses 
• Maximizes revenue 
• Is agreed upon and supported by senior management
From a practical sense, no one service level would fit all situations affecting how long callers will wait. Consider the factors of caller tolerance. How motivated are callers to reach you? What is the availability of substitutes for calling you? What is your competition’s service level? What are your callers’ expectations based on their past experiences? How much time do they have? What are the conditions at the locations from where they are calling?

Five Approaches to Determining the Contact Center’s Service Level Objective

There are essentially five approaches you can use to determine your service level objective, though all require some subjectivity and judgment. 
Alternatives for Choosing a Service Level Objective
• Follow the crowd 
• Relate to competition 
• Minimize abandonment 
• Conduct a customer survey 
• Combined approach
One is to choose a “middle-of-the-road” service level objective, such as 80 percent answered in 20 seconds. The 80/20 objective was once published in ACD manuals as an “industry standard.” In reality, it never was, but many early call centers used this target. 80/20 is still fairly common because for many call centers it is a reasonable balance between callers’ expectations and the practicality of having enough staff to meet the objective. But it may or may not be right for you.
Another popular method for choosing a service level objective is to benchmark competitors or organizations similar to yours, or to use industry surveys that relay what others are doing. Whatever the approach, keep in mind that the results reported by others and what they are actually achieving may be two very different things. I once worked with three different insurance companies with the same stated service level objective — 80 percent answered in 30 seconds. But the results they were achieving were very different.
A third approach is to choose a service level objective by essentially asking, how low can you go without losing callers? This assumes that a higher level of service means lower abandonment and vice versa. One big flaw with this approach is that it assumes that as long as callers don’t abandon, service is acceptable. But that is not always the case. [A]bandonment is not static. It will fluctuate as the seven factors of caller tolerance change. As a result, abandonment is difficult to forecast, and choosing a service level around abandonment is building on the proverbial foundation of “shifting sand.”
Incremental revenue analysis is a variation of this approach and is a more formal methodology to determine the potential impact of abandonment on overall costs. This approach has been traditionally applied in revenue-generating environments (such as reservation centers and catalog companies) where calls have a measurable value. It’s much tougher to use in call centers where the value of calls is difficult to measure, such as customer service centers and help desks.
To use this approach, you attach a cost to abandoned calls and make assumptions around how many calls you would lose for various service levels. The theory is you should continue to add agents and trunks as long as they produce positive incremental (marginal, additional) revenue (value) after paying for their own costs.
Incremental revenue analysis can be valuable when used in conjunction with other approaches, as long as the assumptions are understood and communicated to others in the budgeting process. Nevertheless, don’t let the scientific look of this approach be misleading — it requires some pretty serious guesswork.

A fourth method for choosing service level is to conduct a customer survey. This involves analyzing the seven factors of caller tolerance.

While it’s always a good idea to know what your callers expect, random call arrival means that different callers have different experiences with your call center. Even for a relatively modest service level such as 80 percent answered in 60 seconds, more than half of the callers will get an immediate answer. Some, though, will wait in queue for three to five minutes (assuming no overflow or other contingency). As a result, many in that set of callers would say that your service level is great, while a handful would tell you that it is poor.

Some managers use a variation of a typical customer survey. They have taken samples of individual callers and then compared the responses to actual wait times of those calls. The results are interesting. In many customer service environments, waits of up to 60 to 90 seconds are generally OK with callers. But beyond about 90 seconds, callers’ view of reality can become rather skewed, and their moods can sour. Those who wait three minutes in queue may say they waited four or five. Those who wait five minutes will often tell you they waited something like eight or 10 minutes. Of course, answers will vary based on the type of organization and the seven factors affecting caller tolerance. But the issue remains: At some point, perception deteriorates beyond reality.

Finally, remember that it’s not just how high your objectives are, but how consistently you hit them. If your service level objective is 90/20, but you base your performance on daily or monthly summaries, you might be getting walloped mid-mornings when a lot of your customers are calling. If you really want to see how you’re doing and what realistic targets might be, produce some service level graphs.

A fifth — and overall best — approach to choosing service level is an iterative process that combines all of these methods. See where you are, run some calculations, look at what others are doing, and assess what callers are saying and how they are reacting. In that sense, choosing a service level happens further down the line in the planning process. You need a forecast in order to calculate staff, schedules, etc. So this step initially has to happen in parallel with others.

Whichever combination of methods you choose, you will have more success managing your call center just by having a service level target on which to base your planning. Showing senior management what kind of service can be bought for a specific amount of funding is an excellent way to involve them in this decision and to get their buy-in from the beginning.

Topics: Metrics, People Management, Learning & Development, Site Operations


More from ICMI editors


Leave a comment

Please sign in to leave a comment. If you don't have an account you can register for free here.

Forgot username or password?



Do you currently offer 4 by 10-hour shifts?

Yes, and employees who work this schedule get 3 consecutive days off
Yes, but employees working this schedule aren't guaranteed 3 consecutive days off
Yes, but we give employees working this schedule days off that match customer demand/volume
Yes, but we are experiencing productivity and quality concerns due to fatigue
No, we don’t offer this schedule option
More Polls