··· Chatra All books
§§ Table of Contents − − − − − − − − −

How to measure success

Measuring the success of live chat is simple in some ways and tricky in others. While determining which metrics you’d like to track and compiling the related data is fairly straightforward, judging an agent’s particular chat style or conversational skills is more subjective and can be more difficult to clearly document and explain.

To measure the success of live chat, start with metrics. They allow you to set benchmarks and track progress toward goals. For example, when an agent is new to chat, it can be helpful to let them know what’s expected of them in terms of number of chats per day or how many chats they can handle at once at the beginning, then at 60 days into their employment, 90 days in, and so on.

In general, it’s best to make sure your metrics fit within the guidelines of SMART Metrics. SMART stands for Specific, Measurable, Accurate, Realistic, and Timely.

Let’s take a look at what each of those means:

  • Specific. Choose metrics that directly relate to the processes that take place within the team your creating them for. For example, if you’re seeking to improve how quickly a chat is replied to initially, set a goal — like within one minute — then compare it to the reality and see where they stand.
  • Measurable. It’s essential that every metric is based on actual numbers. If you can’t measure it specifically and in detail, it won’t be nearly as helpful as it needs to be. Make sure every number used for tracking goals and comparison reflects the reality of your business, not an estimate. Use your tools for compiling and providing the data. For example, if you need to know the percentage of customers who used live chat support in the past month, grab the total number of customers and divide it by the number of customers who used chat in the timeframe.
  • Accurate. Accuracy is critical. If metrics are going to be used to make business decisions or track employee progress, they must be accurate. For example, comparing data on how quickly a chat is resolved with customer satisfaction immediately after the corresponding chat can help determine how important a role speed plays in your customers’ happiness with their live chat experience. You can certainly guess that customers are happiest when chats stay under five minutes, but what if customers are really happy at the same rates even when a chat extends to fifteen minutes? Pushing your agents to finish every chat in five minutes or less would be pointless if the goal is to improve customer satisfaction.
  • Realistic. When possible choose hard data and facts overestimates and assumptions. While asking agents for their thoughts and details around the work can be helpful, when it comes to metrics, carefully tracked data is more reliable and useful. A reliable metric is one that is clearly defined and communicated, comes from raw data, and is reported on for every related group or situation in the same way. For example, rather than asking agents what their average resolve time is for live chat, use your chat tool to tell you the facts. People will estimate things differently, so the only reliable measure is with the real numbers.
  • Timely. Metrics should be used for benchmarks and regular improvement. When it comes to support, and live chat especially, metrics from recent interactions — rather than chats that occurred six months ago — are much more useful than older pieces of data. With timely metrics you can determine current agent training needs, make accurate staffing decisions, and generally have a clearer picture of how live chat is going for both your agents and your customers.

Choosing Your Goals and Metrics

When deciding on which metrics to track, consider these three questions:

  • Why are you reporting it?
  • Who are you reporting it to?
  • What outcome are you hoping to achieve?

Each of these questions is key when selecting your metrics at any level — individual, team, or company-wide. Let’s take a look at them:

Why are you reporting?

Make sure the point of the report isn’t the report itself. Collecting data just to do it is a waste of time and won’t be useful to the business. Understand why you’re tracking a specific metric and how it can be used to make an impact.

Who are you reporting it to?

Consider who the metric is for and report it in a way that will make sense to them. For example, a support ops lead may want to know the rate at which customers use live chat, while a product person will be more interested in a metric like how many people have adopted the new feature they added last month.

What outcome are you hoping to achieve?

What’s the point? Make sure any metrics you track and report on have a purpose. While there are plenty of numbers you can present to impress people at a glance, the metrics are impressive for the long haul are the ones that truly reflect the work and a team’s progress at reaching their goals. Choosing certain metrics can also drive specific behavior — so identify what impact this metric will have on what your team does to achieve it.

Metrics to Track for Live Chat

Several metrics work across several areas of support, so you may be familiar with them even if you haven’t started using them for chat yet. As you decide what kind of data would be useful to your team and when making business decisions, your chat tool will help you compile all sorts of information.

Here are a bunch of metrics you may want to track:

  • Volume. How many chats are incoming?
  • First Reply Time. How long does it take for a customer to receive an initial response to their first live chat message?
  • Resolution Time. How long does it take your team to resolve an issue and end an interaction?
  • Average Handle Time. How long does it take an agent to send a message?
  • First Contact Resolution Rate. How often do your agents resolve a customer issue after only one customer chat?
  • Responses per Conversation. How many replies does an agent send during a chat?
  • Agent Utilization Rate. What’s the percentage of time an agent or all agents are spending active in live chat?
  • Response Time. How quickly do agents respond in a chat conversation?
  • Number of Simultaneous Chats. How many chats can an agent manage at once?
  • Proactive Chat Invitation Acceptance Rate. If your team starts proactive chats, how often do customers accept them?
  • Conversion Rate. How often does a chat result in a sale?
  • Peak Hour Traffic. When are your customers most active in your product?
  • Missed Chats. How many chats are missed and sent elsewhere when chat is open?
  • Product Knowledge. While this one has no mathematical way to calculate it, comparing conversion rates, having agents review each others’ work, and looking at customer feedback after a support interaction can help determine product knowledge of any agent.
  • Customer Satisfaction. How does the customer feel after a live chat?
  • Customer Effort Score. How does the customer rate their experience with a company on a five-point scale ranging from “very difficult to “very easy.”

For a deep dive into these and even more metrics you may want to track, check out our User’s Guide to Customer Support Metrics.