Answering repetitive questions from customers with a chatbot (Corona-Usecase)

You can’t concentrate on your core business because your customers or website visitors keep asking the same repetitive questions?

Or you don’t have time to deal with complex customer inquiries because you spend most of your time dealing with simple, repetitive questions?

My tax advisor, Marc Bochsler, felt the same way during the coronavirus crisis. In addition to my calls, he received countless other inquiries by e-mail and telephone about working from home, short-time working, protection against dismissal and much more.

We both quickly realized that this “problem” is an enormously good chatbot use case.

In just 3 days, we had the chatbot online to answer questions about employment law and coronavirus.

Due to time pressure, my client and I opted for a very simple, quick and perhaps even incomplete process. But which processes are still running normally during the coronavirus crisis?

And this is how we proceeded:

1. short voting telephone (60 min.)

Most important questions:

• Tonality of the chatbot
• Appearance of the chatbot
• Name of the bot
• Procedure for content creation

Outcome:

• The bot uses the first name form because Marc Bochsler is also on a first-name basis with most of his customers.
• The bot adheres to the CI of the website.
• The chatbot is initially called “Marc’s deputy”.
• Marc Bochsler, who has been receiving the same questions for two weeks, creates a Word document with all the important questions and answers. I will then structure them and adapt the language to the tone of the chatbot.

2. doing

• I create an account for Marc Bochsler on the Swiss chatbot platform aiaibot.com.
• I create a new chatbot project on aiaibot.com and adapt it to the CI of the tax consultant.
• I start by greeting the chatbot.
• Marc Bochsler documents all questions relating to coronavirus and labor law, writes the appropriate answers and sends them to me.
• I structure the questions thematically.
• I integrate all questions and answers into the chatbot story and follow the tonality from the first voting call (see above).
• I check all questions and answers again, add emojis and further links.

3. testing

• An external proofreader proofreads all dialog for spelling and punctuation.
• Two members of the target group test the chatbot for its tonality and completeness. It is important that the bot covers all important questions.
• Marc Bochsler checks the chatbot again to make sure the content is correct. It is important that I have not integrated any errors when adapting the questions and answers.

4. adjustments

• As expected, minor adjustments will be made after testing. I did this quickly.

5. online jobs

• Thanks to the simplicity of the chatbot software aiaibot.com, I only have to make 2 settings in the software tool and enter the corresponding code on Marc Bochsler’s WordPress website and the new chatbot goes online.

If you have read my other blog posts, then you know how important a well thought-out chatbot concept is to me and how important my self-developed chatbot canvas is for the realization of successful chatbot concepts.

But there are always exceptions.

This special situation required a chatbot with the most important questions and answers about corona within a very short time. So the need was quite obvious. Furthermore, due to the time pressure, too much consideration could not be given to a detailed definition of the chatbot personality, but the chatbot should simply be online quickly.

What happens now?

Of course, the steps omitted above will now follow. The chatbot is tested and optimized with different names, greetings and farewells. Further content is added continuously and existing content must be updated immediately when the law is amended. The chatbot “lives” and “learns”, so to speak. And so are we!

If you have any further questions about this chatbot project, please feel free to contact me at any time. Click here to go to my contact page.

And if you have any questions about coronavirus and employment law, ask Marc Bochsler’s chatbot. You can meet him here.