Emerging Technologies and Markets, User Experience
8 Ways to Create a Chatbot with Great UX
Entrepreneur, Growth Hacker and Marketer, Syed Balkhi says that, “If you want to grow your business, adding a chatbot to your website is key. Think of all the messaging services people use on a daily basis; it’s the preferred way to communicate for many of us.” And it’s certainly true that chatbots are becoming a bigger and bigger part of daily life for most people, particularly in terms of the way that they consume and communicate with businesses. However, whilst Syed is right that a lot of us would rather have a quick online chat to get our problem with a product or service solved, unfortunately that “quick online chat” is less common than we hope! We’ve all had an awful experience with a chatbot, right? Impersonal message that feel rigid, endless “sorry I don’t understand” messages and no way of finding out how to talk to a real-life representative of the company are, sadly, agonisingly bad user experiences that a lot of us have been though. So, whilst chatbots definitely have the power to improve customer service and user experience, they still have a long way to go before they reach their full potential in most cases.

Why is good UX so important for chatbots?

Short answer: Because incredible UX is important in all products and services if you want to keep your customers happy which, of course, you should.
Related article: What Does Good UX Look Like in 2019?
The longer answer though is in part related to the fact that chatbots are replacing human interaction. As such, to be accepted they need to be incredible to use. In general, when used as a form of customer service, a lot of people feel as though bots are less helpful than their human counterparts. This alongside the negative experiences that a lot of us have had means that chatbots need to be working extra hard for good UX.

Chatbots can give a better customer experience with good UX

As with all things UX, their are no cut and dried rules that make it “good” or “bad”. What’s good UX for one chatbot, may be awful UX for another entirely depending on the product or service that they are used within or for. However, there are certainly a lot of general rules that most chatbots could do with taking note of in order to improve the experience that their customers have.

Personalisation is key

Nobody wants to feel like just another customer. They want to be valued and, ideally, all experiences with a brand should give them the feeling that they are. That’s why personalisation is so key within chatbots, particular in the customer service industry. Of course, a bot doesn’t care what your name is or have emotional intelligence, but it’s important that our customers feel as though it does. After all, it’s about mimicking human interaction. Personalisation can be as simple as referring to the user by their name, or a little more complicated like utilising a user’s stated preferences and past purchases to give them personalised recommendations.

Give customers some guidance

Whilst developers and consumers alike can dream of one day being able to have a conversation with a chatbot that is completely seamless, that is simply not (yet!) our reality. To reduce chance of ambiguity and those “sorry I don’t understand” messages we were talking about, chatbots should give users a push in the right direction. For example, through providing buttons to click in response to questions from the bot or by giving examples of how and what they should type. In fact, one study by Raluca Budiu found that the use of buttons improved the simplicity and usability of chatbots:
“Users complained when a bot did not allow them to pick an option and instead required them to type. For example, a participant was annoyed with the Booking.com bot because it did not allow him to pick a date and he had to type it in. After typing in “thanksgiving” he said “It should be something obvious to prepopulate it. Options should be prepopulated — it’s tedious to do it this way. I could hit a button.””

Clarity for the user

Chatbots should make it clear from the very beginning of a conversation that they are just that: Not human! Not only will this build a level of trust with your user, but it will mean that they are more likely to communicate with it in the ways most likely to give them their answers: They may be more aware that the bot may not understand slang, for example.

Get those timings right

Again, we’re imitating human interaction here, so timing is oh-so-important. If your bot responds in lightening speeds, the interaction will feel, well, robotic. However, if it responds slowly, the user will feel like they’ve been abandoned. In her brilliant article, 11 More Best UX Practices for Building ChatbotsEunji “Jinny” Seo says that,
“The wait times should of course, depend on the length of the messages. What we didn’t know until we actually built Labsie was that the wait times also depend on the order of the messages.
When messages are sent right after one another, it feels more natural to pause a little longer and longer with each new message, even if they may be around the same length, so that users have time to finish reading older messages before newer messages are sent out.”

Preempt human error

Unlike bots, humans can’t be programmed to give the perfect response every time! Sometimes we’re ambiguous and sometimes we make spelling mistakes. If your bot can continue with helpful conversation regardless of this, you are on to a winner. As well as using NLN (Natural Language Processing), a simple option can be including a spellcheck system within your bot so that any mistakes are corrected before they are sent.

Test, test and test again!

Before you unleash your chatbot into the big wide world, make sure that you have thoroughly tested it. Re-create various scenarios that the bot is likely to deal with. It’s at this point that you will notice any glaring errors that need to be ironed out. On top of that, analyse the data that you collect during conversation once the bot is live, so that you can see an reoccurring problems, such as points at which the conversation ends abruptly.

Always offer an alternative

Sometimes your bot won’t work exactly how you want it to and sometimes people just don’t like bots! Always provide your users with the option of speaking to a real person, be that via email or on the telephone. At this point also emphasise why your business utilises chatbot technology and how it benefits the user: Dose it reduce the wait for customer service? Is it available 24/7 unlike a call centre? Or is it great for quick queries that people don’t want to make a phone call over? Prove to them that they are speaking to a bot because you value them and their time so much.

User research

In order to create the best user experience possible, you need to first know what your users actually want!
Related Article: Supercharge your Strategy: Bringing User Research + Market Research Together.
That’s where user research comes in. This is a topic that deserves a post of its own, so keep your eyes peeled for our blog post about user research for chatbots coming soon. However, below we discuss an example of how user research can be helpful when it comes to creating a chatbot.

What we’ve learned about UX and chatbots at Snap Out

In the work that we’ve done on chatbots, we’ve found that user research is key in ensuring that the chatbots meet user needs. Through such research, we found that people use chatbots in a similar way to other digital tools, in that they expected a conversational user interface. As Merver Postalcioglu writes for Toptal:
“CUI is a new wave of human-computer interaction where the medium changes from graphical elements (buttons and links) to human-like conversation (emotions and natural language).
Conversational user interfaces are often chatbots — software that mimics a conversation with a real person via text or voice. Chatbots can be used for product recommendations, to book a ticket, order food, and so on. Instead of tapping buttons on a screen to find an answer, users can simply ask the bot a question, like “What’s the latest news on the NBA?” or, “How much money do I have in my account?”…
Conversational interfaces allow companies to create rapid, helpful customer interactions (often more so than with an app or website) and many companies have been quick to adopt chatbots.”
In terms of interacting with chatbots, we found that people used them by either asking questions, as they would to voice assistants such as Alexa or Siri, or through using key words, as they would in a search engine. Therefore, when developing chatbots to have good UX and UI design, we found looking at how people interact with chatbots a key part of user research. This accompanied by looking at the problem that the user was trying to solve, which impacted on how they used the chatbot, paved the way for great UX. The insights helped the development team in terms of the technical architecture of the chatbot, as well as preferences and phrases that needed to be coded in, as it was based on evidence rather than the assumptions of the software development team. We ensured that we tested the chatbots with users to explore their usability and identify whether we had got the technical architecture of the chatbots right. We also tested the chatbot in a real-life scenario to see whether the chatbots met user needs.
Related Article: 6 Reasons That Businesses Don’t Think They Need User Research
As we said, every chatbot is different, but we hope that this article has given you a push in the right direction! The Snap Out Team ?