How to Design An Intuitive Conversational Interface Experience: 6 Useful Tips

Parent Feed: 

It's no longer a matter of “Why?”, but a matter of “How?”. The conversational interface trend isn't going anywhere but... UP.

So, if you want to keep up with the break-neck pace at which your competitors will be launching their own chat-based apps this year, you first need to:

put an end to your “to go or to go conversational?” dilemma. You only have one option...

Then, to start focusing on more pressing issues:
 

  1. what are the user experience challenges to expect when designing conversational user interfaces (CUIs)?
     
  2. what are the best practices to adopt for creating state of art conversational experiences?
     
  3. how do you design a... conversation, after all? One that's natural, seamless, quick and ideally convenient for the user?
     

And once you've pondered on these 3 questions, go ahead and turn the following “enlightening” answers into your trump cards for designing your own rival to Siri, Alexa and or OK Google.
 

1. Put Together An Exhaustive List of Hypothetical Use Cases 

In other words: leave no dead ends in your chat-based app's script. No weak links that could impact your conversation's fluidity.

And how do you get “unpredictability” down to zero? By getting knee-deep in user research work. 

Invest plenty of time and teamwork in identifying all the possible use cases that your chatbot's scrip should cover... impeccably!

How will users react to hypothetical scenario A? And what queries is he/she likely to input in scenario B? What “tricky” request will he/she make?

Once you have THE list at hand, filling in your script with copy will sure feel just like putting the puzzle pieces together. 
 

2. The Onboarding Experience: Be Sure to Give It Special Attention 

The more straightforward your chatbot's onboarding message is, the more chances there are for the user to interact with your conversational interface.

In other words: take no chances, don't try to play smart, resist the temptation to turn your app into a “guessing game”!

Let the user know which are your bot's functionality and key features and how precisely it may serve him/her via:
 

  • a speech bubble
  • an intro message
  • a suggestive animation
     

… from the very beginning.
 

3. Focus on User Convenience When Designing Your Conversational Interface

And convenience in conversational UX translates into minimal user input!

Now here are 2 best practices to consider for achieving the proper level of user convenience:
 

  1. keep your copy crystal-clear and straight to the point: avoid ambiguity at all costs, like giving more than one answer to the same question or answers completely out-of-the-conversational-context, irrelevant to the user's intent
     
  2. keep the user's input to a minimum: the less your customer will need to type in/interact with your chatbot, the better; aim for “yes/no “questions and avoid overburdening them with too many choices to scan through and to select from
     

Extra tip: make your content ideally easy to squeeze through; keep your sentences conveniently short, thus anticipating that the user might head straight to the call to action.

Why should you discourage him/her with an almost impenetrable block of text to keep on scrolling through?
 

4. Set Up Visual Cues-Based Replies to Your Users' Queries

How would you improve readability in “conventional”/non-conversational design? You'd instantly reach out to your not-so-secret trump card — the whitespace — right? 

Well, in case of a conversational design, your trump cards are the visual cues.

Here are 2 examples of how/when you can use them for boosting readability and turning your conversational interface into a truly intuitive AI experience:
 

  1. Use them to assist your customers when in need; for conveying bite-sized information and addressing your users' queries in a highly intuitive way. Maps, charts, infographics images, GIFs will be far easier to “digest” than chunks of text and far more efficient for relaying the message
     
  2. Use them to convey... emotions: whenever words are not enough to convey the feelings you need to convey (e.g. giving them a warm welcome into the app), go with... visual cues; and emoji make the handiest of them all
     

5. Empathy Is Key to Designing a Human-Like Chatbot

A natural conversational interface experience is a human-like one. And a “humanized” AI app is one that easily passes the empathy test. 

Now here are 2 suggestions on how to “inject” empathy into your chatbot:
 

  1. Read the script out loud: how does it sound? Would you say those lines to someone close to you? Do they sound natural enough? Or rather robot-like? Is the conversation fluid enough?
     
  2. Pay attention to context and adapt your script accordingly. Now you wouldn't want your app to give a robotic-like, impersonal reply such as “I hate to hear that” to a user writing: “I feel a bit dizzy”. It goes without saying that you should anticipate all use cases and user queries and have all the context-suitable replies ready (popping up a map showing the nearby hospitals in this case, for instance)
     

6. Embrace the “Less Typing on the User's Side” Principle

For the less they'll need to type, the more fluid the whole conversational UX is going to be.

And, in this case, by far the most popular example to inspire you is Linkedin messager's automated replies. See how it provides users with plenty of choices, right at hand (multiple buttons, “attach a file” option, automated replies etc.)? 

In other words: whenever suitable, automate... everything and aim for user convenience at its best!
 

These are the “tips and tricks” that you may want to consider if you're determined to create a highly intuitive conversational interface experience for your chatbot's users. Ponder on them and see how you can adapt them to your own app's and use cases' scenario.
 

RSS Tags: 
Original Post: 

About Drupal Sun

Drupal Sun is an Evolving Web project. It allows you to:

  • Do full-text search on all the articles in Drupal Planet (thanks to Apache Solr)
  • Facet based on tags, author, or feed
  • Flip through articles quickly (with j/k or arrow keys) to find what you're interested in
  • View the entire article text inline, or in the context of the site where it was created

See the blog post at Evolving Web

Evolving Web