Chatbots Succeed Because Users’ Computer Skills Are Worse Than You Think

My favourite blog post of 2016 was this one by Jakob Nielsen: The Distribution of Users’ Computer Skills: Worse Than You Think.

He detailed a comprehensive OECD study of 216,000 people that showed that just 5% of adults have “strong” computer skills.

This is the chart for the U.S., but it basically looks the same for Canada, the UK and other rich countries

It’s sobering enough to learn that two-thirds of users have poor, terrible or non-existent computer skills. However, what I found most interesting is what constituted a “strong” computer user. Here’s an example of a task that only a “strong” computer user can perform:

One of the difficult tasks was to schedule a meeting room in a scheduling application, using information contained in several email messages.

My clients are managers in internal communications and the digital workplace. When I read that task, I realize that all of my clients, past and present, are part of this elite 5% of computer users. And as strong computer users, we often deploy enterprise systems that look like the one below, assuming that employees can actually use them:

If we trust the results of this study, the truth is that the majority of employees are unable to use the above interface effectively. And that brings us back around to the point of this post:

Chatbots will succeed because people can actually use them.

Everyone understands the chat interface — that’s why messaging is the #1 use of mobile phones. The speech bubbles make sense. You know who’s talking. It’s chronological. You can swipe up for a full history. It looks the same from one chat app to the next. If you know how to talk, or you know how to type, you know how to use a chatbot.

There are, of course, many scenarios that require a graphical user interface. I don’t imagine spreadsheets, video-editing or architectural software will ever be replaced by chatbots. But think about the huge range of scenarios that can be most efficiently communicated with words:

  • “How many clients do we have?”
  • “Turn on my out-of-office auto-reply”
  • “What’s the postal code for our Toronto office?”
  • “Give me the link for Peoplesoft”
  • “What’s the wifi passcode?”
  • “Show me the number for Facilities”
  • “Where can I get a new Skype headset?”

In their Top Strategic Predictions for 2018 and Beyond, Gartner says that “chatbots will become the face of AI”. And I believe it: if an employee need can be communicated with words, it will, eventually, be fulfilled via chatbot.

And then we’ll finally have software applications that everyone can use.

Join me for What We Learned from Deploying a Chatbot for Employees at a Global Insurance Company, a live webinar on January 10, 2018.