Call Center Metrics

[chrissc]

I work in a Financial Institution Call Center, I recently took over as manager and I need some help redifining our metrics. Currently we are using avg speed of answer, acd calls, avg abandon rate, and service level. I want to develop some measurements that look at quality of service as well as these numbers. I currently have 15 people in my department taking an average daily call volume of 775 calls/day. Any help would be greatly appreciatate.

[n_meghanath]

Hi Chrissc

Talking about Service level I don't think there is anything like industry standard however 85% is considered fair enough by every industry.

We assumed that service level is about calls answered within a threshold. Not sure if this would make sense to you however i feel we could talk in two different way. % of calls answered and time factor, For example lets take these scenario :

Call offered : 100
Threshold : 20 s
Call answered: 75
Call abandoned: 5 (within threshold)


1st : (75+5 =80%calls answered /abandoned in 20sec
2nd: 75% calls answered in 20 sec with 5% abandon in 30 sec

which one would you prefer? telling SL as 80% or 75%?
another question would be what about those 25 calls which were unable to reach the agents within 20sec, were they answered after 30sec? 50sec? or 100 sec?

we might considered defining our abandoned No. first, whether to consider short abandoned, abandoned due to IVR help offered etc. then concentrate more towards analysing those defined abandoned which are taken into consideration when calculating SL. It is also important to analyse the wait time for those calls answerred after threshold.

Now coming to ASA (Average speed of answer) it is acceptable only for estimations and trending because it doesn't exactly represent the speed of anwere for all the calls, for instance we may take an example from one of the sample which i went through. 12 agents taking 80 calls per hour with AHT of 7 min can deliver an ASA of 50 sec. However this 50 sec ASA applies to only 78% of the Calls, 22% will generally experience longer wait time may end up as abnadoned call.

The best way is to see how well each matrix fits in to the whole targets, this could be done by establishing a relations between all the matrix, how it affects each other etc. for eg. taking the above example and changing only the AHT to 8 min will change the 78% to a bigger figure and if 78% is what you aim for then it will be hard to acheive.

Let me know if you have a better idea

Cheers

Meghanath
9810607103

[n_meghanath]

oops!!

Not 5% abandoned in 30 sec it is suppose to be 5% abandoned in 20 sec.

[OPsMan]

Hi there,

Those are good measures to provide a quanitiative outlook on your organization. I also use calls per hour for me group. I talk their stafftime and subtract all the idle and available and divide by the number of calls handled. This gives me a true outlook to see what people are really producing.

Are you looking at schedule adherence at all? Antoher metric to consider is AHT? This will allow you to understand where you center is and what possible places you can improve. Any improvment in call handle will give a direct result in your required staffing levels which in turn saves you money.

I would encourage you to go to bluepumpkin.com and participate in some of their free seminars on metrics. They are helpful in educating you on long term strategies for metrics.

Hope this helped.

[JoeB]

This is an interesting discussion, highlighting the many different opinions about the interpretation and meaning of the many call center metrics available to managers. In our research and work with call center managers we find that many struggle with conflicting performance indicators (e.g. improving ASA hurts AHT which, in turn, increases abandonment). A more challenging issue is that most call center can successfully measure and manage internal efficiencies, such as AHT, schedule adherence and occupancy rate, but pay too little attention to customer-facing metrics and, in particular, to leading indicators (customer satisfaction and abandonment being trailing indicators).

There is no simple cookie-cutter definition of “best” measurement criteria. In our Metrics That Matter workshop (www.DiagnosticStrategies.com/mtm.pdf) we discuss how to use a balanced scorecard approach to defining metrics that best represent your organization’s strategic objectives and service level requirements, and to select a small and manageable number of key performance indicators that make day to day management and reporting efficient and effective.

[felix]

If you are not an outsourcer, you might also want to consider other metrics as occupancy.
AHT and service level should give you a good overview of the whole picture.

Working with these 3 indicators you can tell in an accurate way what is the overall caller experience and also how efficient your agents are. (You can get how many calls per hour they are handling, also some other metrics as low occupancy that you might want to avoid the most possible, have your agents logged in when you need them, and perhaps on training when you don't).

Since service level is already giving you a clear indication of the overall customer experience, you might want to add AHT, and occupancy to measure your contact center's efficiency.

Why not measure abandonement rate? If you are running a customer service contact center, if someone wants to call you the odds are if they do not get through the first time they will call again. Therefore, if you are unable to answer some of those calls, the call volume will increase, increasing also abandonement rate if your service level is doing badly.

Service level is a much better indicator and it also englobes either directly or indirectly other measures such as ASA, Aban %, etc.

Service level = X% of calls in Y seconds.

[n_meghanath]

occupancy could be one thing that you might like to consider to check the utilization and productivity of agents. However it is also inversely propotional to the service level.

For example when your service level rise the occupancy rate would generally drop down.

Cheers

[JobGone2India]

Hi,

From personal experience as an advisor who struggled to achieve his targets for call handling time, walk (e.g. form completion etc) et al, I changed my focus to calls per hour and found all the other metrics fell into place and I started to be a star performer (honest!).

With that in mind I have always pushed for CPH to be used in our network as the predominant metric. I recently took over the resource planning side of the network and changed this from an Erlang C method and used a historical calls per hour method which is slightly more accurate in predicting intraday service levels and staff required.

As for our service level measurement we are aiming at:

80% in 20 secs AND 96% in 40 seconds

[ctaylor]

In my experience call center KPI's have varied with the market served, expectations of the customers and what the competition was providing. It is far too simplistic to say 80/20 SL with an 80 FCR and a 4 hour mean time to resolve is what is needed in a technical support center. I have worked with software companies who had an incredible domination of their market (no names please) and they worked to an Average Speed of Answer of 5 minutes at a 70% service level. A computer peripherals client went for 85% of calls answered in 15 seconds with a 90% FCR. Customer Satisfaction is whatever your customer says it is! Ask them whhat they feel they need to be competitive or to secure a competitive advantage. The most prevelant Customer Service metrics are 80/20/10 and 85/20/5, of course there are a number of variation within these ranges. occupancy is another useful metric but if not employed wisely (and reasonably) it will drive turnover. According to 'Cutting Edge Information' turnover rates in financial call centers are as high as 39% per annum. So you must be careful to balance the occupancy with the costs of service and recruitment and training.

If I generalize, which is always dangerous to do, Tech support centers generally have lower KPI standards than do Customer Service centers. We have all gotten use to waiting on hold for a long period of time to get our tech questions answered. That is part of the reason the in some tech centers email now equals voice calls at 47% each!

Chat, FAQ's automated agent responses, IVR as well as email and voice should all be subject to KPI's not just the quantitative types we have been discussing, but also the qualitative ones such as FCR from the customers perspective (it is amazing to me, but there is generally a 20-25% 'gap' between when a 'tech' belives he has solved the problem and when the customer agrees). Other qualitative metrics include Quality scores, versus some form of best practice benchmark and customer satisfaction. According to Tranversal 53% of respondents to a recent survey indicated that their opinion of particular Brands decreased after contacting the company call center! At the same time ASQI has published a report that states that a 1% increase in customer satisfaction equates to a 2% rise in share price.

The customer may not always be right, but they are always the customer and Customer satisfaction is whatever they say it is. Customer Sat should be a heavily weighted KPI in every call center.

If you would like additional information on the sources I've quoted please visit my website at www.thetaylorreachgroup.com and if I can be of further assistance feel free to email me at ctaylor@thetaylorreachgroup.com