Who’s Going To Build Bots?
Many will remember 2016 as the year of botched attempts at conversational interface. The original promise that bots will replace apps had…
Many will remember 2016 as the year of botched attempts at conversational interface. The original promise that bots will replace apps had fizzled. Instead, we’ve seen many bot evangelists back pedal on their original rhetoric. The shiny promise (‘you will be able to chat with businesses the same way you chat with your friends’) got retracted. Now, we see GUI widgets being apologetically reintroduced via the back door. Rather than chatting with a bot, we see pundits now suggesting we tap on buttons and manipulate other GUI widgets.
Why Throw The Towel In?
Funny how people eagerly backtracked on the original excitement. What made them do that? The explanation is simple: it was a spectacular failure of the much ballyhooed onslaught of the chatbots that brought this swift retraction. Recall the tech media at the beginning of the year. The bots are coming! the bots are coming!, shouted the enthusiasts. And then, after the awful fiasco that the first Messenger bots brought, the excitement subsided.
One thing became obvious: we don’t know how to build bots. But how’s that possible? Here we have a pool of amazing teams capable of developing great software. How come they fail when attempting to build bots?
There is only one possible explanation: building bots is not the same as building apps. Just because you have a brilliant team of app builders doesn’t mean that the same team can build bots.
It is such realization that forced people to throw the towel in and to go back to building apps. Yes, they still call those apps ‘chatbots’, but that’s incorrect. There is very little, if any chatting going on with these apps. The new Messenger, Kik, Telegram et al apps are actually good old mobile apps. The only novelty with those apps is that users don’t have to download them. And as big and awesome an advantage it is, still isn’t enough to warrant calling them ‘bots’.
How Are Bots Different From Apps?
There is a fundamental difference between bots and apps. Failing to grasp that difference results in wasting everyone’s time. Talented teams waste time trying to build bots while using the app building skills. Such approach will never work.
So far, all we know is how to build apps. When we build apps, we start by identifying a problem domain. We then focus on building a utility, or a tool that will assist users in solving identified problems. For example, we know that people would like to be able to edit their photos. And we also know that people may lack utilities to do that. Next thing you know, someone builds an app such as Photoshop, and voila! Users can now manipulate their photos using graphical controls that the app offers.
We see that apps focus on enabling users to get things done. Write a document, send an email, calculate their monthly payments on a loan, book a trip, and so on. If we look a bit closer in how do apps work, we’ll notice that they rely on users doing all the legwork. Apps put users in the driver’s seat. Apps enable users to do all the heavy lifting required for achieving their goal.
Apps building teams always assume that app users are the experts.
Bots are different. Bots are the exact opposite of apps. Bots are not tools and utilities expert users will manipulate to achieve their goals. Bots do not assume that users are the experts. Bots are the experts!
Challenges Ahead
As you can see, we currently lack the skills for building such expert systems (i.e. bots). Any time I talk with developers, it is obvious that all they bring to the table is the knowledge needed for building apps. Which is pretty much useless in the bot world. So we need to regroup and rethink our approach to building bots.
Not only from the technical standpoint. We must rethink the game also from the design, user interaction, marketing and analytics standpoints. So buckle up, and join the exhilarating ride!