Average handle time (AHT) is a metric for the total average duration of a single contact, including any customer hold time (if Chat or Phone), talk time, and the follow-up or admin after-work tasks related to that contact.
For chat, you also need to include concurrency in your calculation for AHT by dividing single session times by total logged in "productive time engaged in chats," and all offline/load must be removed from the equation.
There is no doubt that the good old Average handle time (AHT) metric is essential in workforce management workload forecasting. In fact, any type of moving average metric is beneficial for forecasting long-term trends. However, be aware there are some flaws in using averages in any forecasting effort, and much depends on how spread your outliers are in the histogram of the average. For example, as was written by Sam Savage for the Harvard Business Review:
Consider the case of the statistician who drowns while fording a river that he calculates is, on average, three feet deep. If he were alive to tell the tale, he would expound on the “flaw of averages,” which states, simply, that plans based on assumptions about average conditions usually go wrong. This basic but almost always unseen flaw shows up everywhere in business, distorting accounts, undermining forecasts, and dooming apparently well-considered projects to disappointing results.
If you want to learn more about the flaws of averages, here is a link to that article: https://hbr.org/2002/11/the-flaw-of-averages.
Back to AHT. As Cameron Turner says in the article "Let's Talk About Average Handle Time. For a while now, it's been popular in the contact center scene; it's been popular to hate on Average Handle Time as a metric. This thought process, while well-meaning, is, I am afraid flawed.
The advice usually comes in the form of it being the "wrong" metric to focus on because it creates lousy agent behaviors not great for customer experience. i.e., it focuses the agent to concentrate on how quickly they can get through the call instead of how well they can meet the customers' needs or the customer experience.
In other words, this advice is basically saying; low AHT = a Low Customer Experience or, conversely, High AHT = a High Customer Experience. Sadly, however, this is typically bad advice. Anyone who has attempted to correlate AHT with Quality, NPS, or Customer Experience (CX) will find some Agents with low AHT having excellent quality/CX. Conversely, some have terrible Quality/CX or vice versa; some agents have high AHT and excellent quality/CX and some have awful.
I have run similar correlation exercises as Cameron and found there is simply very little correlation – yes, AHT has to be managed in the right way, and agents should not be taking shortcuts to reduce their AHT. Additionally, when assessing an Agent's performance on AHT, it should also be recognized that the average is simply a midpoint of all the conversations that are taking place in that channel. Therefore, when managing AHT, a more useful measure is the normal range of contact and assess abnormally long or short contacts to see why they are different – either short or long outliers are both terrible for forecasting and indicators of something going wrong.
Anyway, what I am steadfast on is that completely dumping AHT as a metric is highly flawed for any contact center - both from a productivity perspective but also as a customer experience outcome – yes, believe it or not, customers don't want long conversations.
Last note on AHT: there is no industry standard for the length of AHT, and it will highly depend on:
The complexity of conversation between the customer and agent and or process the agent has to complete to fulfill the customer's request. For example, a simple "can you send me this document request" could be a very quick AHT vs. a customer complaint. As organizations continue to roll out more digital, chatbot, and app self-service functionality, this will reduce the number of simple transactional conversations, so don't be surprised when you see your AHT rise in this scenario.
The Tenure and Experience of the Agent – this is natural when you first learn a new skill; it takes longer to complete; over time, an agent will become more efficient.
The technology stability and page load time used by the agent.
Check out the weWFM Podcast on Apple or Spotify
Spotify: https://spoti.fi/3J5gsJh
Apple: https://apple.co/3HskI58
Comments