We often talk about UX design in how people physically interact with a product, site or service, but what about how they communicate with it? Conversational UX is yet another example of how design is changing and how us designers must adapt.
Chatbots help people with everything from answering basic troubleshooting questions to recommending products. If you’ve ever interacted with a website’s chat in the last few years, you’ve likely talked with a chatbot. While these are usually less advanced than voice assistants, like Amazon Alexa and Google Assistant, users still want a great experience.
Serving A Much Needed Purpose
We can all agree that long customer service wait times are terrible. You call for support and end up on hold for hours. The same happens with live chat. There just simply aren’t enough support representatives to handle everything.
Conversational UX has filled a much-needed niche – instant customer service. While some chatbots are just for fun, others are equipped to handle common customer needs. From going through troubleshooting steps to helping people pay bills online, they can take care of quite a few issues without an actual person stepping in. Of course, they can transfer you to a person if you need more assistance.
From this standpoint, who wouldn’t love bots? Yet, if they’re poorly designed, waiting 30 minutes to an hour might be preferable.
People Actually Enjoy Them
In one survey, 38% of people said they had a positive experience with chatbots while only 11% said they had a negative experience. While 38% might sound low, some people have never actually chatted with a bot or it was so well designed they didn’t realize they weren’t talking to a live person.
The one thing people are loving most is getting instant service. Anything that’s easy to use and makes a situation more convenient is always a winner.
When designing conversational UX, think about how you would want to be talked to. Do you want to feel like it’s a robot or a friendly customer service representative? Getting the balance between bot and human is tricky, but well worth the effort.
Ensuring Bots Continue To Learn
One of the key elements of conversational UX is deciding what type of chatbot to create. We have to truly think about the purpose of the bot first. Going too complex might make a bot unusable, but keeping it too simple might not address enough user questions and comments.
It’s a common misconception that chatbots magically learn on their own once they’re created. This is only partially true. Much like teaching a human child, chatbots need guidance.
There are multiple self-learning strategies to incorporate based on your bot’s purpose. One common theme is most require at least some level of human interaction, such as adding a new user intent or filling in obvious knowledge gaps. Of course, fixing incorrect information also requires a programmer to step in.
This is why it’s important for us as UX designers to provide ways for users to offer feedback. Users get to provide real-time feedback to help the bot learn and improve the overall experience.
When Frustration Kicks In
Conversational UX doesn’t always work out the way we plan. For instance, Ubisend points out that making a bot too picky ruins the experience. In their example, a user tries to order shoes, but unless they enter the details in a very specific format, which required all caps and dashes, the bot kept saying it didn’t understand.
Despite how simple a command might be, users won’t always follow things exactly. After all, they’re just chatting with someone, whether it’s a bot or a human, and they won’t differentiate between the two.
Users who don’t get the response they’re searching for or keep getting errors will just give up.
Ubisend also makes a very valid point – there isn’t such a thing as perfect conversational UX. The technology is quite there yet, but we can make the experience as close to human as possible.
Mainly, make the conversation as smooth as possible. Also, get to the point. If people actually wanted to just chat, they’d use a novelty bot that plays games or is just for general conversation.
The Best Conversational UX Examples
We can go through all the ways bots are great and not so great, but the best way to tell the difference is by trying out some of the best conversational UX examples around.
IMPACT’s list of 9 Most Innovative Chatbot Examples From 2019 is a must read. We could all learn quite a bit from those bots. Botsociety lists five more great examples too.
Conversational UX is still evolving and the best way to make it better is to think about how you would want to interact.
Image: Andy Kelly & NESA by Makers