“I like the fact that we are moving forward in respect of AI and we are quite admirably keeping up with the others.”

Interview with the chatbot: O2BrainBot and its first mission

O2 Czech Republic passed a significant milestone at the end of 2018. It launched a long-awaited Simple Online Company (SOC) project. However, the system launch was preceded by long preparations for a gradual transition to the new CRM. And mluvii was there.

To get the O2 employees as prepared for the migration as possible, they were trained under the O2 Learning Programme. And because O2 is a modern, digital company that is not afraid of technological challenges, it has decided to use a chatbot to evaluate the training. In this case, O2BrainBot replaced the web forms that the company had used to collect its employees’ feedback. O2 also first used machine learning technology in live communication with users.

Preparation phase

The main task of the implementation team, composed of employees of mluvii and O2 Czech Republic, was the definition of the administrative structure of the questions, the technical settings and the implementation of the chat widget with the chatbot. An important aspect was the automation of the entire communication. The O2 team defined the key questions of the entire inquiry and mluvii then helped with the right contextualization to make the communication with the bot as natural as possible. The process was set up and implemented within two business weeks.

Progress

The chatbot collected feedback from the employees of O2 Czech Republic during the second and third November week of 2018. A total of 2,646 employees of the company were contacted to evaluate the SOC, including 289 managers and senior executives and 2,357 employees. Four questions were defined for the group of managers, seven questions were given to other employees. Completing the questionnaires was not mandatory. In addition to conducting a questionnaire interview with users, the main task of the bot was to collect user satisfaction assessments and their direct feedback in the form of verbal assessment.

Evaluation

Reporting and analytical tools are an integral part of the mluvii platform. And this time as well, they helped with evaluation of the process and success of the entire project. About 900 employees of O2 Czech Republic participated in the SOC evaluation through the chatbot. The return on chatbot interaction stabilized at 25% on average. This means that a quarter of all respondents interacted with the chatbot and successfully completed the questionnaire. If we look at the data more closely, we can find that the conversion rate between respondents and completed questionnaires in case of managers was 45%. In case of the employee group, the result was almost half the figure, i.e. 23%.

A very important element was the evaluation of automation of O2 employees’ feedback and the presence of the chatbot itself. At the end of each questionnaire, the user was given the opportunity to rate the service on a scale of 1 to 5 and write their own feedback. From the point of view of the Customer Satisfaction Rating (maximum of 5 points), the questionnaire was evaluated by an average grade of 4.4.

The figures speak clearly, but equally important are the verbal ratings that users have written. These give us the reason to continue developing mluvii. Here are some of them:

“I like the fact that we are moving forward in respect of AI and we are quite admirably keeping up with the others.”

“Definitely a great idea, filling in is more fun.”

“ChatBot is a move forward.”

“Definitely more entertaining than a classic questionnaire, of course, it is somewhat more challenging to create and maintain, but certainly better.”

“Great, funny, fast, new, surprising.”

“I’ve filled in a questionnaire in such form for the first time, it’s an interesting form.”

“Great, I feel like I’m not communicating with a machine, but some kind of a discussion would be fine.”

“Cool.”

“Cool!”

“Cool :-)”

„:-)“