Conversation UI and Chatbots — They’re Great, But Think Clear about Your Service, Users Before Heading There

Recently I’ve been reading quite a few articles on chatbots and Conversation UI (aka having conversation back and forth on screen with a human or a bot). Apparently, they are one of the hottest trends in the technology world. Given the fact that most users do not download even one app vs an increasing amount of apps published in app stores per month, companies and developers are looking for a better way to productize services and ideas. A conversation-based UI has become one of solutions that gets applause and supporters. I agree that this opens up many new opportunities, including voice-based interactions, but also concerned that people will follow the trends without thinking it through.

Led by the Chinese messaging app WeChat, an app where you have access messaging, service, reading and many other things in one place, has inspired many others to do something similar: Facebook Messenger, Telegram, Kik and many other platforms. They are opening up their APIs and welcome businesses and developers to build chatbots to serve their users. This include a variety of industries: retail, customer service, make reservations, book Ubers, even news digests. Startups of this kind are getting crazy funding these days. I feel the bubbles are bubbling right there.

Conversation UI does great advantages, in certain context, also for who’s going to develop it. If you are going to build a chatbot to help your customers on Facebook Messenger, there are plenty of frameworks claiming “no programming needed”, thus have very low threshold for small businesses to connect users in a more direct way. No need to have a team and build an app from scratch. No hassle on all those glossy marketing. The UI can be “minimal”, meaning you may not need much design aside from those chat bubbles. It’s better for users too – now they finally can get all the services they need in one place, Facebook Messenger or what else, without switching from app to app. And those annoying notification entires that lose us will likely be replaced by more friendly “ABCD company reminds you it’s time to pay for your bills”.

However, just because this is the hottest trend doesn’t mean that it’s good for every service. The first thing is not to follow the UI trend blindly, rather, think about 1. What’s the service/product? Who are the users? 2. What works best, effectively and efficient for them.

Is talking with a service by typing back and forth in speaking bubbles the future way to interact? Well, there are definitely scenarios that we should be cautious about going overboard with it. Think about ordering a cake. Typing back and forth vs selecting options of cakes from screens, which is more effective? As designers, we should keep in mind what we can best help users accomplish what they want to accomplish and how.

I do agree, though, the whole conversation UI thingy will help to push the digital tool to be more human friendly, given the increasing use of artificial intelligence (AI) and natural language processing (NLP). AI and NLP are still maturing and generally require lots of training, if you want your service bot to appear “smart”. Once they get more developed, I can see that digital conversations will become more interesting (and possibly handle a lot more different user requests).

Overall, I think it’s a very interesting topic, even though I’d be cautious about all the “hype” around it and avoid designing a conversation UI just for the sake of following the trend.

Leave a Reply

Your email address will not be published. Required fields are marked *