Comments on: Average Speed of Answer – A Leading Indicator of Satisfaction

By: Tim Childress

Tim Childress — Mon, 23 Feb 2009 21:48:32 +0000

Just curious. Your posting indicates Service Level is usually measured every half hour to get statistically meaningful and controllable sampling. If every call is captured and fed into my Service Level measurement, why do I need to perform that measurement every 30 minutes? Why can’t I just do one measurement at the end of the month?

By: Jan Marais

Jan Marais — Wed, 07 Jan 2009 10:51:35 +0000

Please help. During the calculation of ASA is the number of calls answered within a given SL also counted into the calculation of ASA, or are the calls answered within a given SL taken out when calculating ASA ?

By: Average Handle Time - A Good Metric? For Whom? « Extreme Customer Satisfaction

Tue, 02 Dec 2008 22:13:33 +0000

[...] previous posting on Average Speed of Answer (ASA) remains the most popular posting on the blog, so I thought another Contact Center metric may be welcomed by our [...]

By: Rudy Vidal

Rudy Vidal — Sat, 10 May 2008 04:17:42 +0000

Ted,
Thank you for your comments. I think we describing the same coin from 2 different sides.

I don’t see the cause of ASA as poor staffing. I see staffing as the basic level solution to an ASA problem. However, as you pointed out, the HR budget is key; Contact Centers are often in HR constraints. For this reason I usually choose to see the cause of ASA as internal inneficiencies (which may include staffing or forecasting problmes). I believe the answer is in efficiency and procedural elegance, which will allow us to lower ASA with the addition of less people.

I do agree that at any point in time, if there were appropriate staffing, ASA would not be a problem.

Based on the topic of the posting, I mentioned that ASA was an inversely leading indicator of customer Sat, however I didn’t mean to give the impression that it was the most leading indicator.

Although I also prefer service level, I’m not sure that it is a more leading indicator as much as it is a “BETTER” indicator. It is not clear to me which is able to show correlation with the end state of Customer Sat, sooner.

Thanks for taking the time to comment.
I really enjoy your blog !

Rudy V

By: Ted Hopton

Ted Hopton — Thu, 08 May 2008 17:58:49 +0000

I enjoyed reading your post about Average Speed of Answer (ASA). Your point that a high ASA adversely affects efficiency is important, as is the fact that making callers wait longer correlates with lower customer satisfaction — and you are also astute to point out that a low ASA does not by any means assure high customer satisfaction.

I would like to differ with you on a couple of points, however. First, there’s really only one cause of ASA: poor staffing. You correctly note other factors that could be improved in order to improve the center’s ASA — no argument there. However, every center is capable of providing whatever ASA it desires, simply by correctly forecasting and scheduling the appropriate number of staff. If all the rest of your systems are inefficient, then you’ll need a lot more people on the phones, that’s all. So your ASA really reflects your forecasting and scheduling effectiveness (and your staffing budget, of course).

I would also argue that Service Level is a “more leading” indicator of satisfaction than ASA. ASA is, by definition, an average and so is subject to all the failings of averages as metrics. There are so many ways mathematically to end up with a given ASA that it’s not a very accurate picture of your callers’ experience with your center. By measuring what percentage of calls were answered within a threshold, however, you get a clear picture with one number of the experience that a large percentage of your callers had. That’s the advantage of using service level. And then, of course, you can use other measures to dig deeper into the experience of callers who waited longer than your target threshold.

Thanks for bringing attention to the topic of answering calls quickly!