Canadian Marketing Blog

Service Levels Are Not An Indication Of A Level Of Service.

For as long as anyone can remember, call centres have measured themselves through service levels. This mathematical formula of the percentage of calls answered within a certain time frame has been the bedrock of the industry. It has been widely accepted that by managing service levels you are ensuring a consistent and acceptable customer experience, while appropriately managing your resources. This traditional view of service levels, however, is a fallacy.

There are three main reasons why I hold this view:

1. Service levels manage to averages and not deviations to the average.
2. “Buying back” service levels are a common occurrence in the industry.
3. By not including IVR (Interactive Voice Response) time, service levels do not correctly reflect the customer experience.

No one manages to a service level target for each and every call. By the nature of the calculation, that would be impossible. Instead, we manage to service levels within a given period of time (a day, a week, a month). By stretching out our time horizon, we in affect hide wide swings in service levels that we experience all the time. By not factoring in a measurement that accounts for these deviations, we are not accounting for the consistency of service being experienced by our customers.

The problem of deviations to a standard service level is only heightened by the fact that many in the industry regularly “buy back” service levels. This is accomplished by running at service levels well above target to make up for intervals where the target was missed. This practice does not change the experience of those who waited a longer period to be answered, and is not appreciated by those who are answered faster in exchange. This practice only costs a firm additional money to meet an artificial in-house target.

Finally, service levels do not include IVR time. The IVR and the time taken to navigate it, is as much part of the customer experience as is the hold time to be answered after getting through the IVR. We are as responsible for the experience on the IVR as we are for the experience with an agent. We also have as much, if not more control over the IVR experience as we do over the agent experience.

It is essential that as an industry we challenge ourselves to find new and better ways to account for the customer experience within our centers. I invite people to add their comments and help share what they have done to combat the short falls of service level measurements within their organizations.

Authored by Richard Litvack, VP Operations, Citi Cards Canada

Mr. Litvack has explained a common fallacy in service level measurement. Answering 80% of your calls in 20 seconds does not constitute a quality call.

As a former inbound and outbound customer service representative for both Bell Canada and Rogers Commuincations, I had my concerns about the 80/20 rule.

Despite the fact that both companies have made great stides in calculating and evaluating quality call components, the issue of time spent on a call always seemed to supercede a quality experience for the customer.

Time first, quality experience second. Sad, but true.

Don O'Connor

Mr. Litvack,

Thank you for sharing your perspective on call centre service levels and measurement.

My experience is that inquiries / calls for new services are answered promptly and courteously compared to calls made to resolve issues with on-going services. This shows that generally the call centres are more focused on customer acquisition than customer retention. Though you would agree, it is easier and cost-effective to retain and build business with existing customers than new customers. So there is a need to increase focus on retainig cutomers via improved service and this will consequently increase new business through word of mouth.

Thank you,
Faz

Mr. Litvack raises a good point regarding the "inward" focus that can occur in organizations. Until we capture customer feedback regarding their perception of the experience, and truly know what the bulk of our customers consider a good experience, we're only measuring what we think is important, instead of what matters to them.

This becomes especially important as contact centers are being asked to cut their budgets due to the tight economic climate, and may cut into areas that significantly impact customer satisfaction.

I also agree with the issue of managing to averages. Objectives that focus strictly on averages miss the minimum and maximum numbers and how they can affect the overall customer experience. As Mr. Litvack noted, it's easier for contact center management to essentially hide poorer service experiences if everything is measured by averages.

Thank you,

Rachel Wentink
Sr. Director, Product Management
Interactive Intelligence

Mr. Litvack honestly generalizes and to make comments like we are as responsible for IVR as we are the talk time is ridiculous since this piece is out of our control since the customer is in complete control. If you gauge sl from point of entry customers can make mistakes therefore causing a fake sl which would increase overall cost.

SL needs to be set at a realistic number but, length of call does not guarantee good customer service or experience.
A bit of a fluff piece and unfortunately feeds into a growing trend where contact centre managers do not understand how a call centre runs or should be run, managed or measured as reflected by the above comments as well.