Published: August 05, 2016 | Comments
Live chat metrics can be broken down into three types: those that advance operational efficiency, those that advance customer service, and those that are a poor use of time/energy to measure. Let’s set aside that third type and focus on the first two.
Sometimes operational efficiency and customer service live chat metrics overlap (wait time can apply to both, for instance), sometimes the two are mutually exclusive (handle time and customer satisfaction are often—but not always—at odds with one another).
To make this article as easy to read as possible, I’m going to discuss efficiency metrics first and satisfaction metrics second – but I think we all understand that these metrics apply and interact in a number of ways and my simplification for sake of readability should not be taken as a comprehensive explanation of how live chat metrics can be used, what they can represent, etc.
If you want something more thorough, I recommend this article.
For those who already understand the basics of live chat metrics (or, more generally, contact center metrics) and want an easy-to-read explanation on how to get more out of them, read on.
While some channels are still “under consideration” for customer service organizations (such as texting or video chat), live chat has become an expectation. It’s no longer a question of whether you do it but of how you do it, and as such the culture about live chat customer service has shifted from “how to do it” to “how to do it excellently.”
With that in mind, I’d offer the following as key metrics for improving your return on investment with live chat.
Improving Live Chat ROI by Improving Operational Efficiency
The four most important metrics for improving efficiency are: wait time, handle time, average # simultaneous chats/agent, and trends in queue size.
Wait time and handle time are no-brainers. They represent how long it takes a customer service rep to respond to an initial inquiry and how long it takes them to resolve the customer’s inquiry, respectively. Handle time can be a little bit tricky because some requests simply take more time than others, and being too draconian with your reps about handle time can actually hurt your customer service goals.
If you’re curious about where you should set the bar for wait time, here’s some information that we pulled together about average response times.
Like handle time, the average number of simultaneous chats per agent is not something you should enforce so much as something you should use as a guideline. A fully trained, experienced live chat agent should have no trouble maintaining four simultaneous conversations. A totally green rep can probably manage two without incident.
Obviously then to increase your ROI you need to pay attention to what reps are having the easiest time of moving from 2:1 to 4:1 chats, and how you can replicate what’s working for them to develop the rest of your live chat customer service team. This may come down to better proficiency with pre-made messages, to better time management/prioritization, or to knowing when to hand a customer off to a more appropriate rep- it will depend on the specific rep. If you find that all of your reps are struggling, you might implement a pre-chat survey or revise your pre-made messages to help prep your team for what it is your incoming customers are looking for assistance with.
Trends in queue size is one of the most important efficiency data-points because it helps management identify how large a team they’ll need to staff each type of customer inquiry (if you serve multiple geographic regions or offer multiple, different products this is particularly relevant). These trends might apply at the “time of day” level or at the “time of year” level (anyone who’s worked for an e-commerce business during the November-January timeframe knows what this is like).
In either case, trend data about the live chat queue should help inform staffing and scheduling decisions in order to maximize your team’s efficiency in a way that does not compromise the quality of customer service. You can find an example of a customer of ours that managed queue trends to great success here.
Improving Live Chat ROI by Improving Customer Service
The quality of customer service is harder to quantify. The obvious metrics here would be first contact resolution rates and post-chat satisfaction survey feedback. What these two metrics fail to capture on their own though is customers who never engage your live chat agents in the first place.
Accessibility and convenience are hallmarks of excellent customer service – if your customers don’t feel like it’s easy to get a hold of your team, they’ll never enjoy (and hopefully tell others about) the great customer service experience you know you can offer them. With live chat, it’s almost cruel to offer it and not make it as convenient and accessible to your customers as possible.
So how can you measure the accessibility and convenience of your live chat solution? There are four related metrics and one yes/no “checkbox” to help with this.
The checkbox is simple: if no live chat agents are available (either because they are all at capacity or because live chat isn’t being offered on your site at the present moment), do you communicate that clearly and provide an alternative, convenient means for customers to submit their inquiry to you? This is not an ongoing metric that you need to track (though it may be helpful to track the volume of inquiries you get via email or web-form as a result of live chat being unavailable, and adjust live chat staff schedules accordingly)- but it is definitely something you need to check in on periodically to make sure it’s functioning properly.
The four additional metrics are all connected to the same goal: if live chat is up and running, are customers using it?
If you utilize proactive chat on your site (in which you prompt the invitation to chat with customers), check out the bounce rate for pages that proactive chat is active on. A high bounce rate might mean the invitation is occurring too soon, too abrasively, is not compelling enough as a call to action, etc.
Regardless of whether you use proactive chat or a static click-to-chat button, check out what the click rate is for that button (this should be easy to track using Google Analytics if your live chat software provider doesn’t offer that data in their product).
Next, if you use a pre-chat survey, are your customers completing it and entering a chat conversation, or are they dropping off without initiating the chat? If the latter, perhaps the survey has too many fields/questions, or is ugly, or takes too long to load.
Finally, what is the “missed chat” rate—that is, the percent of customers who enter the live chat queue but leave before receiving a response from the customer service team. This is likely related to wait time but other factors could inform it as well.
Tweaking the way your invitation to chat appears and what the pre-chat process looks like can significantly increase how many customers are successfully engaged with customer service, rather than abandoning your site feeling neglected.