Jose Berengueres, author of Sketch Thinking, talks sketching your emotions, visual thinking and designing conversation loops for chatbots
With chatbots’ rising popularity and the opening of new opportunities to interact with customers, chatbots can now help you overturn parking tickets, become your personal stylist and act as financial consultants.
The growing chatbot trend is showing no signs of slowing down as big players such as Google, Amazon and Apple capitalize on the UX trend with their very own versions. In the future will all our online interactions be with chatbots?
This trend has pushed narrative UX to the fore, pulling language out of the shadows and giving it a much needed revival in the world of UX design.
In our Q&A, author of Sketch Thinking Jose Berengueres shares his thoughts on the importance of sketching, chatbots, narrative UX, and design thinking. Sketch thinking is brainstorming enhanced. It’s using images to convey ideas. Berengueres’ book gives readers a sketching vocabulary and aims to help people in design thinking workshops.
Hi Jose! Before we start, could you briefly tell us about yourself and your area of expertise?
I am a Kaggle competitions master and professor in the Computer Science Department at the federal University of United Arab Emirates where I teach Design Thinking, entrepreneurship crash courses and Computer Science. Our lab research is about robots and human computer interaction. We actively collaborate with industry and startups worldwide regarding applied data science.
You have an insightful book, Sketch Thinking, which guides people to pick up the pencil and sketch their ideas. In your view, how important is sketching in the design process?
We started the sketch thinking workshops as a tool to enable engineers to communicate emotion. The idea came after a workshop at Apple in 2015.
We realized that STEM grads are usually taught technical drawing skills but are never taught how to sketch emotions. This hinders communication and what we call the group IQ.
What are the benefits of prototyping with sketches as your foundation?
As a kid I grew up with an MSX computer and later I moved to Tokyo. While in Japan, I realized that the Japanese are very at ease sketching to communicate ideas in business settings. Whereas in the West, the only businessman famous for his penchant for whiteboarding was perhaps Steve Jobs.
By training or by nature, every one of us has a “preferred mode of thinking”: some people like myself are visual, others prefer to listen (audio oriented), a third group prefers to talk aloud to process thoughts (verbal), a fourth group prefers write thoughts into words to function (writing) and a fifth group understand ideas best when they read them as words.
These are known as the five modes of thinking and each of us excels at one of the modes. Steve Jobs was audio oriented and that is why iTunes was so good. Amazon’s Jeff Bezos is a famous reader oriented person (Kindle).
However, as a visual person I feel that the business world has been “hijacked” by word-oriented thinking. And that is what I loved about Japan, that it is okay to use sketches even at a board meeting. Fortunately, recent works by visual thinking advocates such as Dan Roam (Napkin), Alex Osterwalder (The Business Canvas), and Mike Rohdes (Sketchnoting) have very explicitly shown the benefits of visual thinking.
Similarly, since Apple became the most valuable company in the world, we have that other companies (such as SAP, IBM) have started paying more attention to the UX of their products.
The UX department is probably the only “safe haven” for visual thinkers in today’s corporate mayhem. There are many visual thinkers out there locked in word-only offices who could be happier by getting permission to sketch. And every time the communication of an employee improves, the IQ of the company increases.
We’d love to know your thoughts on narrative UX. Could you briefly describe it for us and what role it plays in design?
Narrative UX is a new term coined a few years ago to describe designing chatbot conversations. Think of HBO’s Westworld narratives, where the chief narrative engineers craft new plays and scenarios for the robot to act. It is a wide open field where no one has found the formula of success because the programming tools do not exist yet. It is as though we have computers but the C compiler and the Cascade Style Sheets have not been invented yet.
Chatbots are very popular now. I remember how simple they were in their infancy. What’s your opinion on them and how have they evolved over the years?
As it was said in one of the keynotes during the F8 conference this year, there is a clear business case for chatbots because they can fill the space between a call center and a static website cost effectively (i.e increase sales at almost no cost).
Website ←—————– chatbot ——————> call center.
Would you explain for our readers the process of prototyping a chatbot?
We recently prototyped a medical chatbot for a VC fund in the valley. This was a challenging experiment. An effective chatbot should maximize the UX understood as the likelihood that a customer returns. Today, the most successful chatbots are programmed by hand by UX engineers using unwieldy decision trees.
An interesting feature of a chatbot is that it is virtually UI-less, the user flow is constrained by the boundaries of a chat and that means the UX depends 90% on the words used (humour, content, story flow), 9% on how they are used (cadence, caps and other…), and 1% on the graphical support elements (GSE) such as photos/GIFS. The cadence is an often overlooked element in chat simulation, it is the delay between replies and there is a ton of research that explains how a simple word changes meaning with the delay… like when someone asks something and I reply yes, or … yes or … … yes…
What kind of functionality would you need from a prototyping tool?
My favorite prototyping tool to use is chatfuel because it lets you deploy a chatbot almost in real time so you can A/B test it with your Facebook friends.
What are some challenges you face in designer-developer communication? How do you solve them?
Regarding chatbots, the number one challenge is to educate stakeholders on the chat design. A seasoned VC partner with decades of web and app design experience might have the temptation to approach this as a chatbot design challenge by copy pasting past strategies. This is a typical mistake. The skills to design a chatbot are not the same that are required to design a mobile app. I would rather hire an unproven Hollywood screenwriter than a UX person with three years’ experience in mobile apps.
Another problem is related to language, English is a terrible language to program chatbots because of the grammar and some other unfortunate linguistic characteristics. However, Chinese with its lean grammar is much more forgiving in terms of AI requirements. That’s why we see chatbot adoption rates higher in China (Wechat, webot).
AI and robotics go hand in hand, as well as chatbots; do you foresee any trends within AI that will become more mainstream in the coming years?
Yes, robots as a service is a compelling business model. In the future, the American family’s most expensive purchase will not be the car but the robot. And the key component of the robot will be the chatbot AI which will be the glue between humans and the internet of things.
Whoever develops the first minimally useful chatbot will have a foothold in the personal robots industry for years to come. Right now, Amazon’s Alexa is the best bet.