Statistics

Of course, the use of statistics is important for every Watermelon user.
How many conversations did your chatbot take care of, and what channel was most communicated with? These are all examples of the statistics that you can find in the Watermelon dashboard.

Choose a time window

To start off, choose a period for which you want your dashboard to show statistics. You can choose from a set standard as yesterday, last week, last month, last quarter or an adjusted period. When you choose a time slot such as a week, month or quarter, the dashboard always takes full calendar weeks and months. The statistics reload every 10 minutes. Can't wait that long? Then click on the refresh button at the top right corner.
Schermafbeelding 2020-02-03 om 10.02.17.png 30.08 KB

The tiles

Schermafbeelding 2020-02-03 om 10.01.00.png 120.92 KB


The statistics begin with 8 tiles. They are explained later.

Total amount of conversations:
This is the total amount of conversations that visitors started. Both chatbot conversations and live chat conversations.

Amount of conversations chatbot:
This is the total amount of conversations your chatbot was able to handle without interference of an agent.

Amount of conversations agents:
This is the total amount of conversations with an agent or employee. This can be both conversations the chatbot has redirected and only live chat conversations.
 
Amount of conversations in business hours:
This is the amount of conversations started between the office hours you set. These are therefore determined by the settings of your out of office time. If you have not set this, then this tile is always empty.

Amount of conversations outside of business hours:
This is the amount of conversations started outside the office hours you set. These are therefore determined by the settings of your out of office time. If you have not set this, then this tile is always empty. 

Average speed of pickup time by agents:
This is the average speed at which all agents pick up a conversation. This figure is only measured during office hours. The timer starts when the last message from the visitor comes in until the agent picks up the chat.

Average handling time agents:
This is the average handling time of all agents. It is measured from the first message sent in a conversation up until the last message sent.

Amount of conversations per channel:
This is the total amount of conversations received per channel connected to your dashboard.



The linegraph

Compare the chatbot conversations to the agent conversations. The dark blue line, as you can see in the image below, shows all conversations and therefore adds up the chatbot conversations together with the agent conversations. Ideally, you see the number of chatbot conversations increase and the agent conversations decrease. Because, of course, you want your chatbot to work more and more independently. You can also export the line graph statistics. To do this, click on the export function at the top right.
Schermafbeelding 2020-02-03 om 10.01.14.png 135.64 KB


Tag statistics

You also have the option to view an overview of tags that have been used. How many conversations were about a specific topic or, for example, were successfully handled? You can easily find and review this. Choose Total amount of tags from the drop-down, to see all tags to date. You can also export these tag statistics. To do this, click on the export function at the top right corner.
Schermafbeelding 2020-02-03 om 08.28.38 (1).png 18.81 KB


How can I use the tag statistics even better?
To be able to measure even more specifically based on the statistics, you can use tags. If you add a tag to every conversation, and then archive it you can start to take stock of, for example, the chats with the tag "successfully handled." In this case, you always have an overview of what the chatbot delivers to your organization. For example, you know that the chatbot picked up 250 conversations last month with the tag "successful." If these 250 conversations had all been an email or telephone call, you would have spent about 3 minutes on average per question. In this case, the Chatbot has taken over 750 minutes (over 14 hours) this month from a human employee.
 

Export your conversations

To get a good insight into how conversations proceed, there is an export conversations function in the statistics environment. Here you can export all conversations or only agent conversations. The results are shown in a .csv file.

How do you do this?
Choose a period and click on -export conversations-. Then choose whether you want to export all conversations or only agent conversations. A .csv file is downloaded for the chosen period.
Was this article useful?