Organize Shift Work for Your Live Chat Using Tags

By in LiveChat Blog > Business management,
Optimizing your live chat shift work

It’s good to know the chatting habits of your customers. This knowledge allows you to adapt your chatting schedule and shift work around people if you are getting more chats on a particular part of the week or day.

We like to get into the heads of our website visitors by predicting not only the time of the day when we will get the most chats but also the topics that will be discussed.

See how we use tags and reports to take a peak into the future to prepare our shift work for the incoming chats.

Shift work prediction through tagging

Have you heard about one of our newest functions –tags? It’s a neat little feature that allows you to assign a certain tag to each chat. Using tags, you can group similar chats together and then filter your chat reports and archives according to a specific tag.

Tags have also other uses. For example, we use them to predict when and what chats we’ll be having. We do this by tagging most of our chats (around 85 to 90 percent) with five tags that describe the purpose of each chat.

The five tags we use are:

  1. Support – All manner of support enquiries belong to this category. If a customer is experiencing some problems or needs help setting up a particular feature, the chat is tagged as support.
  2. Sales – Chats from potential customers who are thinking about setting up a trial or are currently testing out LiveChat on their websites.
  3. Confused – We use this tag whenever we get a chat from a person that mistook us for another company or service.
  4. Empty – Our agents mark a chat as empty when there is no real interaction with the visitor. The chat is either completely empty or the visitor turns silent after the initial ‘Hello’.
  5. Feature request – This tag allows us to keep all feature request handy and available for our development team.

Most chats receive one tag only, with rare exceptions where a chat belongs to two categories.

By setting up a list of pre made tags for our agents to use, we are able to keep our tag list clean and concise, without dozens of similar tags cropping up (trust me, using the same tags is really worth it).

Using the tags like this gives us huge amounts of information about the types of chats we are having with customers. After looking at the data we’ve been collecting for the last couple of months (since the introduction of tags), we can tell you a lot about our chats.

For example, we get the most confused chats during the night shift work. Those chats amount to about 35 percent of the chats we get at night. Feature request chats are also the most common during the night (they are still pretty rare though, with about 0,5 percent during the night shift).

Hourly distribution of confused chats

The busiest hours happen over two shifts during the day (10 a.m. – 9 p.m.). This is where we get the bulk of our support and sales chats (35 percent for support and 5 percent for sales). Just before that, we get a lot (even up to 25 percent) of empty chats (7 a.m – 12 p.m.).

Hourly distribution of support and sales chats

Now, this might look a little bit scary, as confused and empty chats combined come up to around 50 percent of all chats. Thankfully this is just a part of the equation. What actually happens is more clear when we also take a look at the duration of those chats.

Empty and confused chats seem to be much shorter than sales and support chats. The median for a confused chat almost reaches the three and a half minutes mark. The median for an empty chats barely goes over one minute.

Our support and sales chats are much longer, with support sitting at a median of almost ten minutes and sales at a ten and a half minutes. All in all, even though we get a lot of the empty and confused chats, they are a small portion of the actual chatting we do at LiveChat.

Median Chat length

Adjust your shift work to the chatting patterns

After you identify the times when you get the most productive chats, you can play a bit with your schedule to reach the point where your support coverage is optimal. You probably want to have more agents when there are a lot of sales or support chats and less agents when the majority of chats is from confused visitors.

Since we knew that most of our valuable chats happen during the two ‘day’ shifts, we focused our attention on them by updating our shit work and delegating additional agents to those hours. During the day, we have three to four agents available. The remaining night shift work is handled by two agents.

What was the result of the change in our shift work? More happy customers. When all the important chats are handled immediately (since there is more than enough people available to handle them), customers are bound to leave good satisfaction rates.

When looking at customer satisfactions statistics, empty and confused chats received much lower rates (66 and 68 percent respectively) than the sales and support chats (95 and 96 percent). Even though the overall satisfaction score averages at 81 percent, we know that the most of the chats that really matter receive nearly perfect scores.

Customer satisfaction of tagged chats

To see how well optimized your support shift work is, there are three steps you need to take:

  • Start tagging chats – Without the data, you won’t really be able to tell when the important chats happen. The tags can be different for your company. However, make sure your agents are using the same set of tags, since unifying them later on will be a never-ending task.
  • Export and interpret the data – Filter your chats by tag and export the results to a CSV file for easy interpretation. Compare the results to the total number of chats to get a clue about the scale of use of a particular tag. Enable the 24-hour distribution option to discover the peak times for particular types of chats.
  • Adjust your shift work – After you determine what kind of chats you are getting and when it happens, change the shift work for your agents to cover the busier areas with more people. This way, you will provide an ample response to the increased demand for answers from your clients, resulting in increased satisfaction rates.

A simple thing like adding a tag to a chat can help you immensely when applied on a larger scale. It’s almost like having your own personal oracle that predicts what kind of chats you will be having. As the time goes, you can add more and more tags to your list to monitor the frequency, satisfaction levels and other metrics of specific types of chats.

It will help you and your time organize your shift work to make sure all customer queries are handled efficiently.

Photo courtesy of Matt Biddulph via Creative Commons.


comments powered by Disqus