AHT as call center management evaluation metric

arrow green upMy article on Productivity resulted in quite some good comments. AHT as a metric is also discussed on other blogs. In this blog I further detail my arguments why I believe AHT is still a powerfull metric to track in Contact Center environments.

Focus on AHT can kill the Customer Experience

In general CSR’s behave as their bosses want them to behave. Meaning: If management’s sole focus is AHT you will see CSR’s do just that. And nothing more than that. You get what you deserve and it kills the Customer Experience.

I believe it is quite “dumb” to ask each and every CSR to have exactly the same AHT. This does not do right to the differences there are between human beings and converstations. Even dumber is to ask from agents that each call takes the same (short) time.

These are all examples and consequences of bad usage of a powerfull metric.

AHT used in context can be a powerful tool

Still I’m pro AHT as a good measurement tool in contact center environments, if used in the right context and with care.

My line of thinking in bullets:

  • High quality means a high consistency in quality (high likelyhood that each call shows same (high) level of service experience
  • If you want to be consistent in high level customer service experience you need to know exactly what you (through your CSR’s) need to deliver and how. You need to have a clear methodology.
  • You get this consistency through intense training and coaching of your CSR’s on the methodology (before that: make sure systems and authorisations are in place to deliver what you want delivered) and guard it through quality monitoring
  • The result is consistency in quality, execution (methodology) and therefore consistency in AHT
  • High AHT and Low AHT of an individual CSR is an indicator of poor quality delivery. Either knowledge or structure in conversation is bad, or CSR is rude and “cuts” calls. Both do not stick to the methodology and will not deliver the service experience you need.

Either way: both (groups of) CSR’s need attention. Not on AHT, but on Quality.

I ensure you: perform a correlation analysis between QM-results and AHT on all CSR data, and you will see what I saw: the correlation is (very) high.

Too many outliers says something about management too

High AHT-variance or -spread proves a lack of management capability to:

  • know what you need to deliver and how to do this most effectively
  • develop a clear methodology for flawless delivery of the desired service experience
  • explain, train and coach CSR’s in understanding and executing the methodology consistently

As an example I include two graphs to show what I mean. These graphs show on the X-ax: AHT-slots from lowest to highest. On the Y-ax: number of CSR’s that “perform” within a certain AHT-slot. The result: you get a nice view on how the spread of AHT is within your Contact Center. In both cases AHT is 285 seconds. You tell me which Contact Center is more likely to deliver consistent quality in service experience:


Bottom line 1: AHT-variance or spread metric is therefor also a management evaluation tool.

Bottom line 2: AHT can never be the #1 or only metric you use. Use it in the right context.

11 thoughts on “AHT as call center management evaluation metric

  1. Pingback: Call Center Managers: Don’t Stop The Small Stuff. | Call Center Cafe

  2. Pingback: Keeping Call Centers Running Efficiently | CallCenterBestPractices.com

  3. Wim, I am pleased to ascertain that other experts, like Tripp Babbitt, in the meantime made it clear that AHT (Average Handle Time) is very oldfashioned (rooted in the “Command and Control thinking”, a productivity mindset of over 100 years old), has completely run it’s course and is no longer tenable in Contactcenters !

    Please go here (Ctrl + click) : http://blog.newsystemsthinking.com/blog/my-brand-of-insanity/0/0/call-center-aht-average-handle-time-wrong-measure-wrong-solutions


    • Hello Vermeer,

      Thanks for coming back. I would like to share my answer to the question of Blake Landau on the CustomermanagmentIQ blog

      Her question was:
      If you don’t allow call center representatives the time to know the customer, how will that call center representative capture feedback so management knows “what” and “how” to deliver?

      My answer:
      I do not recall having said that size matters..

      Seriously: great question that allows me to clarify explain that I really am all in favour of taking time to solve an issue, request, complaint or create an additional sale. Whatever time it takes to get it done right the first time.


      The size of AHT (of the entire call center / skill group etc, measured over a seriously longer period than 1 call) does not matter to me. What matters to me is, that with a certain, given, AHT you need to look out for the “outliers”: the people that significantly deviate in their Average Handling Time (of let’s say a 1000 calls) compared to the overall average.


  4. Pingback: Contact Center Intelligence Sunday Summary week 21 « Contact Center Intelligence

  5. Good and oh so true comment. If you want to compare apples with apples you need to factor in call types.

    From my experience that’s not too difficult if you use skill based routing. You should be comparing employees in the same skill-groups.

    AHT should be a real average and that means you should not measure AHT on a daily level. Weekly is possible, depending on the volume handled. If you take a large enough sample on AHT (let’s say more than 500 or even a 1000 calls) you will not have a negative effect though of uneven spread of call types (under the assumption that you compare within skill-groups).

    Thx for your good addition to my blog.


  6. AHTs best use is in conjunction with the actual reason for the call.
    If you are looking at variances in AHT, and the call center takes a wide variety of call types with varying times, then an agent who randomly gets the ‘tough’ calls will be penalized. Alternately, if the tough and easy calls are distributed evenly, then AHT will have a high variance.
    Only when evaluating AHT against a call type – and comparing that to AHT of other agents for the same call type, can you use AHT as a good evaluation tool. A higher AHT indicates that the agent may need coaching in that particular area. Well, assuming that they don’t have the best Solve Rate (FCR) in for that call type; in which case every one else needs to talk for longer. But that’s a different can of worms.


  7. OK,

    I will post this here and the same comments at the other blog (since I promised I would).

    This is why I think AHT is a flawed metric. It is a hardware-centric, efficiency-driven metric. It is used as a crutch by managers who just want to CTA when dealing with management. If I manage metrics properly, there is no problem. If I spot a problem, I have to solve the measurement problem – not the cause.

    Solving the cause is what drives effectiveness and happier customers. You said above how managers who reward and punish based on AHT get agents that act on that metric. This does not lead to effectiveness and happy customers. It leads to unhappy customers who are cut-off the phone at the precise time.

    I agree with what you are saying up above re: looking for variances in AHT to spot potentially quality problems. Sure, that is one way to use efficiency metrics, make sure it all sticks to averages and things continue to work as always. Except when something changes and the averages are off. If you are just looking at AHT variance you will miss the real reasons why those changes happen. Here is an example.

    Let’s say that MSFT came up with a new patch for their latest OS (yeah, unthinkable – i know). That patch was implemented automatically and “broke” things down. The people who call to complain about it will undoubtedly spend more time on the phone with your CSRs. You will see a variance in the AHT.

    The question you need to ask yourself is “WHY?” not, “WHO IS RUNNING MY AVERAGE UP?”.

    See, this is where AHT does not offer the quality of information you get from looking at other variables. If you were to be monitoring feedback (agents and clients, both), or you had the process in place for agents to contribute to the KB with the solution (if found), and look at feedback (again, the pesky word) you get more visibility into the reason why your variance occurs.

    Would you rather know that there is a variance or find the reason for it?

    This is the problem that most CC have when looking strictly at numbers and statistics. They miss the why and the solution.

    If you can leave AHT behind, you can then focus on solving the issues, ratcheting higher effectiveness and making customers’s happy. Whom, I promise you, are not happy to spend more time on the phone with you and walking away empty-handed.

    It is not about running an efficient CC anymore. Using direct feedback your customers will let you know who is performing and who is not. If you trust them, you can stop looking at statistics and useless metrics from the past.


Comments are closed.