ICMI is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.


13 Key Metrics for Chat Success

Chat as a support channel delivers many benefits to an organization. Since agents can handle many concurrent communications with end users, it supports a lower cost per session than traditional voice communications. At the same time it delivers real-time communication, which has been shown to deliver higher levels of customer satisfaction.

For any company that provides support via chat, achieving these benefits requires an understanding of the metrics that help you monitor the health of your chat channel and, when necessary, make improvements.

To understand the metrics you should be measuring, you first need to understand who interacts with chat. In general, this means focusing on the end user (or customer), support agent, and support administrator. By looking at all three, you can monitor the health of your entire support structure, rather than just a single component.

Key metrics for end users

Your customers want their issues solved quickly, so they are looking for a responsive chat with someone knowledgeable enough to ultimately resolve the problem or query. If the chats are being queued, customers want to know how long they should expect to wait.

Monitor the following metrics in order to ensure that you respond to and resolve customers’ problems as quickly as possible.

1. Average time waiting for chat to be answered. Are customers waiting a long time before they even begin the process of chatting?

2. Length of chat. Once they’re chatting with an agent, is the chat itself taking too long? The right answer depends on the type of support being provided (detailed technical query vs. order status).

3. Agent response speed. When the end-user types something, how does it take to get a response? The number of concurrent chats that the agent is handling affects this; it’s important to get the balance of response speed and agent efficiency right.

4. Time to resolution. How long does it take to resolve the entire issue? Also, what is the percentage of chats that are resolved by the original agent without having to pass the customer around?

Key metrics for agents

When looking at agents, efficiency and effectiveness as paramount. Expectations for agent metrics like how many concurrent customers they can help and how long it takes to resolve an issue will differ by industry and by issue type.

These metrics will help you keep an eye on agent performance so you can make informed decisions about workload management and identify areas for performance improvement.

5. Number of concurrent chats. How many simultaneous chat customers can your agents handle? The more the agent can handle, the lower the cost per session, but this efficiency needs to be balanced with customer perception and satisfaction.

6. Overall chat length. How long does the chat session last? Manage the expectations for the support organisation as a whole and compare against industry averages.

7. Length of post-chat wrap-up. Is the process to close off after a chat too long? Can workflow changes speed up this process? Using forms may help to limit the information required by agents to what is strictly necessary.

8. Average agent availability. Are you best using your agents’ time? If there’s high availability, how can you better use their time; if availability is low then you may be overloading your agents. Understanding this metric over time shows you the peaks and troughs of availability and helps you better predict the right resource levels in the future.

9. Speed of end-user response. Are slow-responding customers draining your agents’ productivity? It’s important to understand this metric and identify trends so an agent is not penalised for overly long chat sessions caused by slow customer activity.

10. Use of automated answers. How frequently are agents able to speed up the overall chat by using canned responses? Is there an option to provide the information elsewhere on the website so the customer does not need to contact support?

11. Individual agent stats. In addition to the overall agent metrics above, it’s important to track a few key metrics on a per-agent basis so you have a measure for comparison and can identify areas for individual agent improvement.

  • Average chat length 
  • Average post-chat wrap-up time 
  • Number of chats offered
  •  Number of chats handled

Key metrics for administrators

The administrator is responsible for helping the team deliver the right level of support at the right time for the best customer service. Monitoring the number of incoming chats, monitoring service level agreements (SLAs), and critically reviewing workflow to look for necessary adaptations are key to keeping things humming along.

The following metrics will help administrators see the big picture.

12. Number agents available vs. number incoming chats. In addition to agent availability (discussed in the section above), what is the ratio of available agents to number of chats? Monitoring this ratio helps you understand the flow of your support system and overall load on the system.

13. Average number of parallel chats. How many chat sessions are happening at any given time? This also helps you understand the overall load on the system. Combining all of these allows for better resource planning for the peaks and troughs of your support center.

Although you can get significantly more granular with your data, this snapshot provides the main metrics you should consider tracking to ensure customer satisfaction, agent effectiveness, and overall support workflow, efficiency, and cost. It’s important to understand the benchmarks for your industry and to set your own goals, so you can always be striving to meet--and beat them.

More Resources