Crafting conversational UI for your chatbot is an important step. Having understood the basics about Conversational UI in our previous post, let us now move on to understanding the steps to be considered while crafting Conversational UIs.
With the speed at which our current technology is developing, it won’t be long before we start conversing with our computers to get things done. In developed nations, this technology might be already existing today. The kind of technology they have access to and because they are at the forefront of the development of any related technology, they have gained an upper-hand. So, let's begin to understand the basic essentials while crafting conversation UI for chatbots.
How to craft conversational UI of a Chatbot?
Statistics at a glance
It is estimated that 95% of consumers believe that the future of customer service is mostly relying on chatbots. With 28% of top performing world businesses already using AI for marketing and the 32% getting ready for AI integration within the next year, it is clear where the future of marketing lies. As if these figures were not already enough, it is estimated about 80% of world businesses will have some sort of AI integration by 2020.
The need for crafting Conversational UI?
Even though we have covered this in our previous article, a conversational UI (CUI) is an interface that allows you to interact with applications by using voice/speech commands. A CUI lets you interact with the various application on human terms, essentially just by talking to them. CUI can be classified under 2 main categories –
As we journey towards the future, many conversation interfaces are being developed. Sadly, a majority of these are generating extremely poor experiences for the user. There are two fundamental problems that needs to be addressed for this issue to be addressed.
Firstly, texting is prone to miscommunication
Meaning, a conversation is a mix of both verbal and non-verbal communication. A chatbot is built to understand context based on conversational flows and speech pattern recognition not based on intonation, facial expression, eye-contact, gestures or posture of the user. In other words, the non verbal communication is not factored into calculations.
Albert Mehrabian conducted a seminal study that shows that, non-verbal communication amounts up to 55% of all human interactions while the content that is actually spoken amounts to only 7% of what is communicated and 38% of interpretation is based on how we speak. Meaning the copy written for chatbots needs to be carefully crafted around the information that you seek/provide.
Secondly, conversational interfaces are still fairly new
Since they are fairly new, there is no set/ universally acceptable method of designing conversational interfaces. The knowledge gained by designing Graphical User Interfaces will not be much help here as they are both entirely different modes of interface.
The experience in a conversational interface and chatbot ui doesn’t depend on the flow, or the icons or the colour that we can see through our eyes, but rather the conversation itself. Hence, crafting and curating good content is the only way to ensure that the user feels engaged.
What are the guidelines for designing a CUI?
Most industries will employ AI Chatbots or some sort of AI integration by 2030, that alone must have you running to get your own chatbot or voice assistant. What will decide how much you benefit from this is how effective is your conversational design. Think about this in terms of a technological shift. Since we have some experience in the chatbot industry, here are the concerns we recommend you consider before embarking on an adventure to find the perfect chatbot for your organisation.
Think about it as the personality and functionality that you want to associate with your Chatbot or voice assistant. This should be a value addition for the user experience. Something that will increase engagement and also address an issue your users have been struggling with.
The process of employing a chatbot or voice assistant needs to be a thought out one that solves real issues being faced by the user. If you already have an engaged user base, address a common concern they are all facing. It will engage your users more and they might also explore more options you provide with the bot. They should be certain that they can depend on your bot for solving said issue in the future.
First, concentrate on solving one issue for the user. Rather than focussing on all issues that a chatbot can solve for you, focus on one issue. It must be the one that you want the interface to solve when you are crafting the UI. There's so much more you can know about chatbots. Here's a complete article on it by Forbes.
We have the greatest upper hand over machines when it comes to reasoning. There are clear indications that there are a few tasks that humans agents handle better while some tasks, on the other hand, that chatbots manage better. Repetitive work, information gathering, lead generation are a few areas where automation can really help you out in cutting costs while maintaining the same guidelines and procedures even when the number of users is scaled up.
Another thing to keep in mind is, fast resolution of issues makes happy customers and they tend to return with more business.Estimations show that in the US alone, 30% of such repetitive tasks currently handled by customer support managers can be automated to save approximately $23 Billion a year.
Just like people, bots can be either funny or just basic. If you have understood the first point we made, you will understand the role that personality plays in this case. Do you want to satisfy customers with your services or do you want to engage them with conversations? Will they return for the speed of service or the conversation? This will contrive by the way you address your conversational interface needs.
A simple AI chatbot with a couple quick witty comebacks for people that like to ask silly questions will be much more suited when you want a straightforward yet engaging bot. If your whole service is based on entertainment, you’d probably require a bot personality that is more enforced with really good and near perfect copy to answer questions.
Knowing your users will definitely prove to be very helpful for deciding the tone and conversation flows that you need to employ for each different user scenario. You need to research the potential audience that you intend to reach. You must have a clear idea as to what the customer expects from you, discover their needs and wants and cater to them. Understanding the problems, fears and motives of your users will help you to come up with creative and innovative solutions adding value to your offerings while you are crafting.
Getting an understanding of the channel through which users stumble upon your bot will be of great help to you. Knowing whether the user comes to you by a referral, by clicking on your link on your blog, via a paid advertisement will help you decode their goals and expectations. Based on these, one can develop relevant conversation flows and maintain context in the conversation.
Chatbots have been here for a while now. Whether your chatbot is capable or not depends on how much you saturate your field. There are some categories and services that are already crawling with 100s of chatbots and some where chatbots haven’t been employed yet. Depending on the amount of competition there is already in your field, you can decide whether to make your bot an intelligent one or one with a great sense of humour.
Your bot can definitely define your business, the how is up to you. The personality of your bot can speak for your whole brand. Make sure you use the best possible language to convey your values through the bot.
Adding a little less of personality may make you look like you are not even trying and adding a lot of personality can make you come off as trying to be someone you are not. Maintaining your social branding overall, in colour, style, font and language will earn your bot some extra points.
To achieve a consistency and give a personality to your chatbot, write down the important characteristics of your bot. It's just like how you create a fictional character with a backstory and personal info. You will also have to pay attention to the name and avatar of your bot. Sharing the personality trait file you made for the bot with all the parties involved in creating it will ensure a continuity and coherency.When dealing with users, always remember to have a great approach and be mindful of where the content can be misinterpreted. For such cases, designing better conversation flows and better copy can be your saviour.
These days, anybody can create a chatbot in a matter of minutes without coding. In other words, even your competitors can create one with just as much ease.The challenge is to create a believable and more engaging chatbot. That takes a lot of good copy and experience. Also, it helps if you have a technical team. It will know what they are doing when it comes to AI integration.
When a user starts a conversation with your chatbot, there are many ways it can branch out. You can plan for all those branches and figure out conversational flows for them all. Or else, you can just keep the user within bounds by using clever content.
If you are going to use a conversational interface, identify the key metrics that will help you gauge the effectiveness of your chatbot. These metrics should resound with the purpose of bot creation.Look at the driving metrics to identify pain-points. Know how engaged your bot-users feel, look at the time they spend on an average conversing with your bot. Boost customer satisfaction by keeping a track of your product returns, measure website users through chatbot to monitor traffic.
In other words, generate and analyse data, use the insights gained into the demographics of your user base. Understand user engagement and retention numbers. Figure out what are the points at which the users drop-off. Check whether there is a common point at which this happens.
Use the data to fix problems while crafting conversational UI. A direct feedback when executed with finesse can help you generate great solutions. The complaints of unsatisfied users often provide great insight for developers.
If you're convinced and would like to experiment more, you can also check out Chatbots Magazine and their amazing insights on conversational UI.
Crafting conversational UI
Wrapping up now. We have tried to give an insight and understanding of building a conversational interface. You now have the knowledge to create your own bot and gain the most of it. Going with a chatbot is a great way to increase your business productivity and reduce costs. While providing continuous support and customer satisfaction, Chatbots are also great tools to collect and analyse data.
For more on crafting conversational UI and chatbot technology, please visit Engati Also, read our blog on Conversational UX
It's time to apply your learnings from this blog by building your first chatbot. Register to get started!
Blog Cover Photo by You X Ventures on Unsplash