You want to start a chatbot project? Then you’ve probably heard of my chatbot Canvas?

I developed the chatbot canvas for the first time in 2019 and since then it has been downloaded over 1000 times and also used by myself in countless chatbot projects.

At the same time, a lot has happened on the chatbot market. Conversational assistants are attracting more and more attention. Technology is constantly evolving and the technical possibilities with chatbots are becoming ever broader.

So it’s time to update my chatbot canvas and that’s why I’ve created version 2.0.

The following questions in particular have been added:

1. What added value does the chatbot bring for companies and users?
2. What are the limits of the chatbot project?
3. What are the reasons for the respective technology decision?
4. How can the planned chatbot be further developed?

And this is how a successful chatbot is created with the Chatbot Canvas 2.0.

1. challenges

Hand on heart – what is currently causing difficulties for your company and your customers? Where do you lose your customers? Where are your customers annoyed by long queues or missing information?

The first step is to think about the current challenges facing your company and your users. Focus primarily on challenges that could be solved by a chatbot.

In many projects, a chatbot primarily solves the following problems:

• Overloaded customer service
• Lack of customer service outside of normal working hours
• Insufficient provision of relevant (customer) information
• Difficulties in gaining new leads
• Lack of incentives to give feedback

However, other problems are also possible. It is important to note that there is no clear right or wrong answer here – every company has its own individual challenges that a chatbot should solve.

2. added value

Imagine you have to justify the chatbot internally and fill out the budget request. What added value does the chatbot bring? Added value can be on the customer and company side.

You often have different use cases in mind at this point. The collection of added values quickly shows which chatbot use cases are most likely to be worth implementing.

I usually recommend creating a list of priorities here. The use cases with the most added value are placed at the top and are implemented first. Other cases may follow in later projects.

3. the chatbot

Here you roughly describe the idea of your chatbot. What should it be able to do? what typical dialogs will he have? You are welcome to sketch an example dialog here or, if you are inspired by chatbots from other companies, you are also welcome to name them here.

4th target group

At this point, I don’t like to read expressions like “our customers”. Rather, it is about who the customers are and what characteristics they have. In the next step, it is important to understand in which situations they use the chatbot and what they want from the chatbot. These questions should be answered here. Take a little more time to think and analyze at this point, it will pay off later.

5. functions

Here you list all the functions and features that your chatbot should be able to do later. This usually starts with the question “Does the chatbot have to understand free text or is a simple rule-based chatbot enough?”. A Natural Language Processing (NLP) component is required to understand free user input.

Other functions can be interfaces to internal systems so that the chatbot can make database queries, for example. Or perhaps the chatbot should be able to create conversation logs? What types of media do you want to play via the chatbot? Should users also be able to upload documents to the chatbot themselves?

And what role does data protection play in your project? Is it important that the data is hosted on-premise or at least in the EU?

This is just a small selection of chatbot functions. You can be creative here.

6. limits

No, the chatbot will most likely not be a jack-of-all-trades. Like every technology (and, incidentally, every employee), a chatbot reaches its limits somewhere. This is not a bad thing, but it should be known from the start.

This field is new in the chatbot canvas and I find it very important. Even though chatbots are becoming increasingly intelligent and, thanks to advances in robotic process automation (RPA), are able to automate more and more processes, there are still things that chatbots cannot do. If you are aware of these “gaps”, you can use other means to try to make these gaps fade into the shadows.

7th technology

Anyone who knows me knows that I’m not a fan of premature technology decisions. That’s why I deliberately put this point in seventh place. It only makes sense to decide on a technology once you know exactly what your use case and requirements are.

I usually recommend choosing a technology that can still be used when the chatbot is further developed.

8. testing

Everyone is committed to it and then the project is behind schedule and testing is skipped. However, it is so important that you test the chatbot in peace before it is released to the masses. So think about how the chatbot should be tested during the project conception phase. Who tests it? What should be tested?

And then plan a little time for possible adjustments.

9. marketing

Not every chatbot needs an additional communication strategy. But most bots do well when they are not just suddenly there, but when there is a communication strategy for the chatbot. This applies not only to internal communication, but also to external communication.

First of all, all employees in the company should be informed about the advantages and limitations of the new conversational agent. Depending on how well your colleagues are already familiar with chatbots, it may also be worth giving a general introduction to the topic of “chatbots”. Customers or potential customers should then be informed about the new channel. This can be done via social media or newsletters, for example. There is also the option of the new digital assistant simply suddenly appearing on the website and contacting the user at the bottom right of the browser window as required.

The communication effort at this point varies depending on the use case and target group.

10. further developments

This point is also new in the canvas and I think it is very important. Most chatbot projects do not end with one project but are part of a larger customer service or digital strategy. It therefore makes sense to consider from the outset how the chatbot can be integrated into the strategies in the long term and, above all, how it can be further developed.

Would you also like to have the chatbot canvas and fill it with your content?

Then simply enter your name and e-mail address here and the chatbot canvas will be sent to your e-mail inbox.