Global - en
Choose your preferred location

The Chatbot Experiment

Interview with Ingo Waclawczyk, Meetup speaker and UX expert

The 20th UX Meetup Metropole Ruhr was held at our Design Center and presented the perfect occasion to discuss a special topic and special experiment: 
“Empathy and compassion for a better UX? Let’s ask the chatbot.” This was the title of the most recent Meetup, which takes place at alternating locations throughout the Ruhr area every two months.We spoke to the Meetup’s Speaker, Ingo Waclawczyk, Senior Manager for User Experience, about the extraordinary chatbot experiment at the Meetup, and about UX in technical dialogs.

For the Meetup, you chose as your speaking point a theory by Paul Bloom from his book Against Empathy: 
Could you explain the theory and tell us what inspired you to choose it as a topic?

Ingo Waclawczyk: Paul Bloom’s position can broadly summarized as follows: Empathy – contrary to what the provocative book title suggests – is not fundamentally bad, but, in some situations, provides a poor compass for decision making. According to Bloom, this is particularly true for decisions that affect groups of people, rather than individuals. 
I first became aware of this thesis when it was presented in a lecture at the Global Peter Drucker Forum 2018 in Vienna. In actual fact, the theory was only touched upon lightly in the lecture given by Rasmus Hougaard. But the subject immediately aroused my interest and started me on a quest for more information. I therefore contacted Rasmus and asked about his sources. It was in this roundabout way that I eventually came across the papers by Paul Bloom and his book Against Empathy. Bloom is a professor of Psychology at Yale University. In his work, he addresses the question of how children and adults perceive their physical and social environment, focusing, among other things, on language development and morality. 
Of course, a provocative title like Against Empathy is also an example of great marketing. Nevertheless, I have studied his theses closely and the experiments and studies on which they are based – and wondered whether and how these findings can be transferred or used in the field of user experience and usability. As UX experts, the design of human-machine interfaces is our daily bread of course. Therefore, the question arose as to whether the latest findings on human-to-human communication could also influence human-to-computer communication.

Speaker Ingo Waclawczyk at the UX Meetup held at the Design Center, Phoenix-See

As a UX expert, have you often noticed that machines communicate in a manner that is too factual and mechanical?

Ingo Waclawczyk: It is apparent that many projects have been – entirely unintentionally – planned, considered, and implemented from more of a technical perspective. This can be down to many factors, including, for example, a traditionally technology-heavy, IT-driven perspective, the existence of established processes, or cultural influences. The fact that a process of communication also takes place with machines – and by this I also mean digital machines such as computers – is usually considered from more of a functional perspective. For me, as a UX expert who develops digital products, it’s all about how exactly communication between people and digital systems works. 
For a simple example of a technology-driven approach, let us take a look at error messages in an online contact form. When a user – who is often using the form for the first and only time – enters something that the system does not recognize or cannot assign, a message such as, “Mandatory field not completed,” will appear; ideally in red and with exclamation marks to underscore the user’s ‘error’. 
From the perspective of the technical system, which requires data to be input in a certain way known only to it, this is certainly a useful statement. However, for the user who has completed the form with the best of intentions, the information is not at all helpful. The message suggests the user has done something wrong and, moreover, does not provide any information or help to clarify what the system does require. This is most likely, therefore, to engender a feeling of having not done something right – something we humans tend to find rather uncomfortable. It is a prime example of a negative user experience: The user feels bad despite having done everything right. 
In fact, the effort required to design the communication process in a way that would engender a positive user experience is minimal. In the above case, it would be sufficient to formulate the error messages to both help the user and avoid making them feel bad.

Empathy and compassion are clearly major factors in human interactions.

Are they similarly instrumental in human-machine interactions (UX)?


In the Meetup, visitors had an opportunity to experiment with three different chatbots, one of which communicated in an impersonal manner, one in an empathetic manner, and the third in a compassionate manner. 
Did the results of the chatbot experiment turn out the way you expected?

Ingo Waclawczyk: Yes and no. For example, we had expected the impersonal chatbot to perform well because it corresponds to what users are familiar with in the real world. What was surprising, however, was that the users arrived at their destination fastest with the empathetic bot, even faster than they did with the supposedly efficient technical bot. Another result that I found remarkable was that more than 90% noticed differences when using the chatbots, noting that it was the most compassionate and sympathetic bot that best guided them through the process. On the other hand, there was a lot for users to read with this bot, which, with a smartphone in particular, was found to be annoying. Compassion takes time, you might say – and empathy gets you there faster. Of course, this cannot hold true universally, but only in certain contexts and situations. We need to look into and clarify this in more detail.

Did you expect such a big audience at the Meetup? And did any more questions come up afterwards in conversation with the visitors?

Ingo Waclawczyk: In the run-up to the UX Meetup, we discussed various topics for the lecture and it quickly became clear that this was a topic that interested and appealed to many people. In particular, the question of actual application and tangible benefits often arose during preparation for the event. In that sense, I was not surprised to see so many participants registering and attending. 
The conversations with the visitors before and after the experiment were also very interesting. ‘Empathy’ was mentioned by everyone I spoke to – but its distinction from compassion was, it appeared to me, just as new to many as it was for me before I saw the lecture in Vienna. 
Before the event, we wondered when we could consider the evening a success. An important aspect was ensuring that the visitors had an enjoyable evening and were able to go away with new impressions and inspiration. From the visitor feedback, it looks like we achieved this.

What other key user experience topics would you like to see covered? And would you be interested in speaking at another of these events at the Design Center?

Ingo Waclawczyk: For me, the most important factor when it comes to user experience of a digital product is usability – in other words, the simplest possible operation for the user and their tasks. If a user cannot complete the task they want or need to do intuitively and without impediment, then they simply will not use the digital product – a contact form for example – or will not use it again. And worse, instead of keeping it to themselves, a user could even share their negative experience with others. This is why I believe that maximum usability is one of the main success factors for digital products and brands. 

Thank you for your time and the fascinating input on the topic of UX!

More on UX to follow in this blog. 


Related articles:

MeetUp Metropole Ruhr