Analytics for Zendesk
Zendesk Analytics Package
The Zendesk Analytics Package provides insight into your customer’s behaviour and how requests to your support system are handled. All overviews are designed to provide at-a-glance value, no complicated setup and tuning is necessary.
To get started using the Zendesk Analytics Package a connection to your Zendesk instance must be created. Enter your Zendesk web address and credentials, and we will start processing the information in your Zendesk instance.
Changes in Zendesk, like new incoming tickets, are continuously processed and added to the analytics package. This is very much a “live” window into your customer support operations, showing how the situation is right now.
There are four main analytic categories: overview, ticket characteristics, sentiment analysis and ticket content.
The overview page provides top-level insight into the state of your Zendesk instance, including elements of ticket characteristics, sentiment analysis and ticket content. By default, the overview shows data from the last week, but the time period is customizable. Analytics about the entire timeframe (from when the Zendesk instance was created and up until today) is therefore available.
Tickets are ordered by which agent they are assigned to and the status of those tickets. This graph shows which agents are most productive and which tickets get solved, as well as the workload for each agent.
It is therefore easy to spot the most and least productive agents. Based on this information, new tickets can also be assigned to agents which have few tickets assigned to them in order to distribute the workload between the agents in an optimal manner.
The sentiment of tickets is calculated based on their content, and they are classified as either Very Negative, Negative, Neutral, Positive or Very Positive. In the screenshot above, the bulk of the incoming tickets are negative. This chart can be used as a gauge as how satisfied your customers are with your products. If the bulk of tickets are neutral or positive, then there is little cause for alarm. But if most tickets are Very Negative, then action should be taken to investigate why customers are so dissatisfied.
The ticket content is analyzed, and the frequency of words and sentence fragments is shown. If certain words are often repeated in incoming support tickets, they are shown here. On the screenshot above, “Damaged package” is a frequent sentence fragment. This might indicate your shipping partner is too rough with the packages you send and might be worth looking into. By analyzing the textual content in incoming support tickets in this way, important trends and problems can be detected, which are otherwise easily missed.
The number of new support tickets that have been received in the last 7 and 30 days is provided as well, plus numbers for solved and closed tickets. With this overview, it is easy to keep track of the volume of support requests and if your support organization is able to handle the volume.
This page dives into certain characteristics about the tickets in your Zendesk instance, like where tickets are created (Facebook, Twitter or email) and the priority of the tickets.
All tickets have a priority, and this graph shows the distribution of tickets across the priorities. It also shows the type of ticket, for example: questions, incidents, problems or tasks. In the graph above, all tickets are incidents.
The priority distribution of tickets can indicate how severe the problems in the tickets are. If there are mostly Low and Normal priorities on tickets, this might indicate a normal state, while a majority of High and Urgent tickets might indicate serious problems with the quality or usability of the products.
In Zendesk, customers and users can be configured to be part of organizations. This is typically a customer company or another logical entity. The overview shown above shows the number of tickets created by members of these organizations, making it easy to see which of your customers needs the most help from your support department.
Customers can get in touch with companies through a wide variety of channels, from Facebook to Twitter and email. But which channel is the most popular? This graph answers that question, and in this case email is by far the most popular channel.
Knowing how customers contact your support department enables you to strengthen that channel. If Facebook is the most popular channel, then posting articles on your Facebook page about the topics most customers ask about or face problems with may help your users find answers themselves without contacting your support department. Knowledge like this helps you target which information to post about your product and where to post it.
Knowing which customers are happy and which are unhappy is crucial information for the support department and company management team. Based on the content of support tickets from customers, we calculate a sentiment value ranging from zero to one, where zero is very angry and one is very happy. This sentiment value enables a ranking of the most positive and negative customers, see the graphs below.
When the most dissatisfied customers are identified, it makes resource allocation decisions much easier. And the positive customers can be singled out for tasks like product ambassadors or similar marketing activities.
The sentiment trend shows the changes in ticket sentiment over time. The graph above shows the change in the last week, showing very little change. In the graph below, we show a much bigger timespan – 5 months. The overall trend here is that the sentiment of tickets is stable.
A downwards trend can easily be spotted in this graph, meaning that incoming tickets are increasingly negative. This means that customers are more and more negative in the language in the tickets they send in, and this is a cause for concern, which can help direct customer support strategy. Similarly, a trend rising over time indicates tickets from customers are increasingly positive.
The content of tickets is analyzed to reveal which words and sentence fragments are the most frequent based on ticket priority, organization and channel.
The analysis of ticket content by priority enables you to see which words and sentence fragment are used in tickets with high, medium and low priority. In the screenshot below, “credit card” is a term used quite frequently in tickets with normal priority. Since this might indicate trouble with using a credit card to pay for your company’s products, this is certainly something which should be looked into.
The ticket content is also analyzed on a per-organization basis. This shows which words and sentence fragments that are most frequently used by this organization, and can be used to pinpoint specific problems in this single organization.
We also analyze ticket content on a per-channel basis, based on from where the tickets were sent. Using this overview, we can see which topics are most widely mentioned in tickets received from Facebook compared with tickets received by email.