You know the story.
You’ve probably read Steve Jobs’ bio or at least seen some of the movies about him. You’ve most probably owned an Apple product or know somebody who does. You’ve surely been involved in the iOS vs. Android debate.
It was thanks to Apple that people in tech started thinking about the importance of design and user experience. Steve Jobs’ vision of end-to-end control over what’s happening left no room for mistakes. People had no control over the process, but they had the functionality they needed. Even better — by removing the hesitation, Steve channelled the whole experience and made it straightforward.
We all know that. Today everybody working in the UX and UI use this knowledge when developing apps and websites. We’ve mastered the graphical user interface (GUI). But we are entering a new era now — the conversational commerce as Uber’s Chris Messina named it.
It’s shiny, promising and hyped. But more than that, compared to the GUI the conversation user interface (CUI) addresses a much more complex problem — chaos. We no longer have a visual interface that we can navigate around. We need to think outside of the box, but with the graphical interface now gone, there simply is no box to navigate around.
Mastering the chaos
Making design intuitive was a hard task and one that took many years to master. Making conversations intuitive is like landing on Mars.
Of course it’s possible, just not yet.
Apple saw that and back in 1988 a project was under development. It was called the Knowledge Navigator and was supposed to be the very first artificial personal assistant. As visible today — the tool never made it to a commercial release. Actually, nearly 30 years later we still can’t reach that level of cooperation between a human and a machine. Not with a portable device for sure.
Nevertheless, we are slowly finding our way in the chaos. We try to apply with chatbots, what we have learned from GUI. We create conversational flows to guide the user, we place buttons to ease the choice and we build our chatbots with a single narrow purpose. Or we don’t. We embrace freedom of speech and try to build smarter bots instead, not limiting our users in their expressions.
What’s the right way?
The Zen of Steve Jobs, by Caleb Melby
Questions bot makers need to ask
How should we design our bots? Are we building a feature or a product? Can we make a business out of it?
Chatbots are like any other product. There are two main questions we need to ask:
- Who’s it for?
- Why should they buy it?
Once we get that straight, each one of us need to honestly see where they stand:
- Can I provide the best Natural Language Processing? Yes / No
- Can I compete with companies with millions in funding and hundreds of engineers? Yes / No
- Well, who do I compete with, after all?
Competing with the human curiosity
The driving force behind the human race is curiosity. So it’s something really big and endless. It is The Force. Can a team build a tool that can answer all the world’s curiosity? Can we build an unlimited interaction with limited resources?
“Alexa and Siri are so wide range, they have problems with NLP.” — Matt Marshal, Venture Beat
Chatbots use a lot of NLP, but as Stefan Kojouharov notes, most of them work with predefined question and answer scenarios. It’s not just easier and more accessible, it also provides a better experience for the user. Artificial intelligence is surely advancing, but it cannot produce thoughts on its own. While we work towards that goal, the safe way for the business is to train the monkey to repeat what we taught it.
What becomes more important now is what we teach our bots. We are moving from the technical to the human aspect.
The best thing a bot can do is to be straightforward
“Bots are not a technical problem. Bots are an editorial problem.” — Robert Hoffer, SmarterChild
To provide a great UX, bot developers need to focus on the chat bubble. Text (speech) is the new interface. No one likes to read long messy texts, many don’t have the time or patience to read a good one. We live in an immediate world and chatbots need to support that reality. The best thing a bot can do is to be straightforward. This will not just remove clutter and lower user hesitation, but also set the expectations early on and guide the conversation accordingly. A lean interface means a rich experience.
“If I had more time I would have written you a shorter letter.”
— Pascal, Cicero, Twain, Churchill and many others
The ability to express one’s thoughts is becoming popular again. Dale Carnegie’s books are not getting out of style any time soon. Words matter, and they matter even more when you can’t take them back.
New types of jobs will emerge, demanding for well-expressed and emotionally intelligent people. Their role will be to develop conversation scenarios and to make bot interactions efficient, while at the same time — pleasant. Chatbots may not have emotions, but as they deal with humans, they will need to take that in to account.
Journalists, marketers, psychologists, writers. These are some of the people today who can teach bots how to be understanding, articulate and genuine. They are the story tellers, people who can convey a meaning with fewer words, the right words.