ChatBots are undoubtedly one of the most talked topics by Digital Marketing experts. The reason of this success is clear. Thanks to advanced AI systems and their ability of emulating engaging conversations, ChatBots can now provide assistance, release news and information, activate services requested by users and drive customers along the purchase process.
Incorporating a ChatBot into your Marketing strategy means changing the way you interact with your audience. It has also a lot of benefits in terms of costs and it is available 24/7.
ChatBots have the potential to automate a lot of activities and autonomously held a large number of conversations; after all they are software programs and for this reason, it is necessary to monitor several key metrics that will tell you if you’re taking the right road.
Here there is a list of 10 metrics to consider for delivering the best Customer Experience with your ChatBot:
- Total number of users reached: in marketing, percentages matter; to verify the validity of all your metrics and nurture your ChatBot knowledge base it is necessary to have a large number of users.
- Average session duration: reaching a reasonable number of users is not all. The session duration between human and ChatBot has to be satisfactory. The average session duration depends on the kind of services provided by the company and the ChatBot activities: if they have to answer to users’ requests, the average session duration will be shorter; if they have to place orders or tell a story, the interaction will be longer.
- Errors and malfunctions: by constantly monitoring ChatBot-human conversations, you can easily recognise issues occurred and also when your ChatBot didn’t provide any answer.
- Number of sessions per user: this metric is strictly related to the kind of ChatBot: if it has been created to sell products and services, a large number of sessions indicate that the user is engaged. On the contrary, if the ChatBot has to give information and support, a large number of sessions could indicate its little ability to answer properly and satisfy users’ requests.
- Click-through rates (CTR): it is the ratio showing how often people who see your ChatBot, then certainly interact with it. It can be used to measure the performance of your Virtual Assistant. ChatBots will become important indicators, when sales and support services will be accomplished solely through them.
- Active users vs Engaged users: there is a big difference between an active and an engaged user. The first ones only receive and read messages sent by your ChatBot, whereas the second ones are more engaged while they interact with your Virtual Assistant. In order to understand how to improve the user experience and their level of engagement, it is necessary to monitor both categories.
- Complex requests and difficult conversations: it is important to monitor the way your ChatBot works when it has to handle non-linear conversations, by measuring how many times it shows confusion triggers and analysing each case individually in order to prevent mistakes from happening in the future.
- Response time: ChatBots have to give instant responses; in fact, their success depends on how fast they are able to answer. For this reason, it is necessary to test them constantly and interact with them, in order to understand if they are working correctly.
- Conversation steps: if the aim of your ChatBot is to create emotion, engage people and deepen the interaction with them, the conversation should contain more steps. If your Chatbot has to give direct and straight-to-the-point responses, the conversation should have as little steps as possible.
- Life cycle of each interaction: this indicator depends on the purpose of your ChatBot too, i.e. if it has to help users place orders, all information must be retained until the day their order is received; if it provides daily updates, it must be set up with an established retention rate of information.
How many tools are necessary to monitor and analyze all these metrics? The answer is ADA. It allows you to create autonomously your own Virtual Assistant, without any coding knowledge. With the ADA platform, it is possible to measure the conversation potential, the user level of engagement and interaction, by extracting data and information from both Facebook Messenger ChatBots and those integrated into your website.
Detailed dashboards show the number of conversations, the average depth of answer, the ChatBot knowledge base, the number of questions with no answer, the geographic distribution of users, the browser they have used, FAQs and so on.
ADA analytics represent a powerful marketing tool; Marketers can take advantage of the ChatBot potential to deliver the best Customer Experience.