A complete guide to chat AHT, and how to optimise it without adding resource

AHT, or average handling time, is a metric used in customer service spheres to get an idea of how long customer interactions typically take. Chat AHT, then, measures the duration of live chat conversations.

Naturally, you want to make sure that chats are happening as efficiently and productively as possible. But you don’t need to add extra resources to do that.

Indeed, you can optimise chat AHT using the functionality already baked into your chat channel. Here’s how.


First, a note re chat AHT

Chat AHT is the average time it takes for a live chat agent to handle an interaction. It spans the duration of the active chat conversation.

So, unlike call AHT, it doesn’t include any time that the customer spends waiting to connect. With chat, queue times are a metric typically tracked separately.

Nor does chat AHT include any time spent in post-chat activities. (Such as the customer filling out a post-chat feedback survey, or the agent adding chat tags, for instance.) With chat solutions, such activities are not typically time-tracked as standard.

So, chat AHT spans the time an agent and an operator are chatting – from the time the agent picks up the chat to the time either party clicks ‘end chat’.


The chat AHT formula

To calculate chat AHT, you:

  • 1. Add up the total time spent against all chats within a given reporting period
  • 2. Divide that by the number of chats an agent opened within that period

So, for example:

  • • Agent X opens chat #1042 and spends 10 minutes on it
  • • Agent X opens chat #1043 and spends 5 minutes on it

Agent X’s total time spent on chat is 15 minutes. During that time, Agent X opened 2 chats. Thus, Agent X’s average handle time is 7.5 minutes.


AHTs across channels

AHT will vary considerably depending on the channel in which the interaction took place.

Email, for instance, has longer AHTs than the telephone. It’s not real-time, and so expected reply times differ. Support tickets take longer yet. And live chat sessions have the shortest AHT of all.

Or rather, they should have. As a real-time messaging channel, live chat AHT should be rapid. But this efficiency varies from contact centre to contact centre; from chat deployment to chat deployment.

Based on a study conducted with our own WhosOn customers, contact centres following a strategic chat implementation plan have substantially improved efficiency than contact centres without one.

For example:

“Customer A”: Implemented WhosOn without a chat implementation plan
“Customer B”: Implemented WhosOn with a chat implementation plan

A side-by-side efficiency comparison can be seen below:

As you can see, efficiency metrics vary – and chat AHT is just as subject to change. So, how can you optimise chat AHT, without adding resource?

The answer lies in smart use of live chat features.


1.      Proper chat routing

Setting up robust, skills-based routing is a great way to reduce chat AHT. Skills-based routing means that every chat will go to the best-placed person — the agent with the knowledge and resources to answer the query in question.  

As a result, fewer chats get extended handling times caused by inefficient transfers. Agents don’t waste time on queries they can’t handle. And customers don’t have to spend excessive time waiting for the right person to get their query on their screen.

2.      FAQ bot

Deploying a chatbot equipped to handle frequently asked questions is another optimising chat AHT strategy.

An FAQ bot answers your common, routine questions before (or if) an agent needs to step in. This optimises your chat AHT metric in two ways.

First, chats are shorter, as agents don’t need to spend time copy-pasting FAQ answers. Second, your AHT is more representative of the handling times for the complex and important interactions, and not skewed by the simple stuff.

3.      Canned responses and auto text

The crux of optimising chat AHT is reducing it by creating more efficiency. Canned responses and auto-text are two great features that help with this.

Canned responses are pre-written answers to commonly asked questions and conversation topics. They reduce the keystrokes needed for an agent to supply the right, on-brand answers.

Auto-text is what allows agents to access the right canned response for the situation. It stops them from having to scroll through pre-written messages.

Both these features enable agents to respond quickly and efficiently. They don’t waste time typing common or predictable things. This reduces handling time and creates a clearer picture of how long it’s taking to solve queries, not write out pleasantries and instructions.

4.      Pre chat surveys

Pre-chat surveys help to make sure agents have all the contextual information they need. This includes the basics like name and needed department, and can also cover specific needs.

So, instead of ‘I need payment help’, for instance, the customer has already supplied more information by being prompted to give it. Agents get ‘this part of my card isn’t working’ or ‘I cannot get to the payment page’ before the chat has even begun.

This extra information saves the agent time. In turn, it streamlines chat AHT — the agent doesn’t get stuck doing investigative work into vague queries.

It’s also possible to use a bot in lieu of a pre-chat survey. For the customer, this makes the experience of providing extra preliminary information more ‘conversational’ in nature.

5.      Agent empowerment

Empowering agents is a must for all customer service teams. That is, agents need to have the power to help a customer and put right grievances themselves. This empowerment improves chat AHT across the board. After all, there’s no need to escalate, or transfer, or wait for permission to solve a query. (All of which take unnecessary time.)

See also: Agent empowerment: what is it, and why is it so important?

6.      Co-browsing

One thing that can hinder great chat AHT scores is when chatters and agents need to try to explain what they are seeing and doing. (For instance, if a chatter is having trouble navigating the website.)

In such a situation, live chat co-browsing is ideal. With it, agents can see what the customer is doing directly. They can then advise accurately and interactively. Rather than, that is, trying to understand a potentially obscure explanation of the problem, which will inevitably take more mental effort and agent time.

7.      CRM/database integration

Another hurdle to optimised chat AHT scores is poor tech. In particular, a lack of integration with the CRM or required databases.

CRM integration into your live chat channel means that the customer information your agents need is readily available. So, agents face no issues searching for and retrieving information. They don’t have to battle with channel or window hopping.

In turn, all their focus can go on efficiently and effectively providing great experiences.


Optimising chat AHT: a final note

Average handling time is an essential metric for any contact centre, across any contact channel. But when measuring AHT, it’s important to remember that efficiency and quality don’t always go hand in hand.

Optimising your chat AHT isn’t just about reducing the time a chat takes. (Though, admittedly, it plays a big role.) Live chat features undoubtedly help with speeding and streamlining conversations – but features alone don’t make for an optimal chat support session.

Rather, to optimise AHT, you also need to optimise the experiences you’re creating and measuring.


Useful links

Meeting the need for customer service speed

Chatbots vs webforms

For all their training, your agents are only as good as their tech