Over the course of our month of blog posts focused completely on chatbots, you’ve probably gathered that we think they’re here to stay.
With its market value set to rise to $1.34 billion by 2024, and with 77% of customers having improved perceptions of a business after talking with them online, there’s no doubt in our minds that chatbots are on the up.
Related article: The future is bright for chatbots
So, with more businesses choosing to use chatbots, it’s also more and more necessary that people know how to create good ones that leave their customers happy, instead of frustrated!
At the heart of this is brilliant user research.
However, like just about every emerging technology, it can be tricky to know how to implement the findings of your user research within the design process. Here are our top tips.
Find out more about Snap Out’s user and market research services here.
Utilising user research whilst designing your chatbot
Understand how your customers talk to your brand
Conducting user research interviews is the ideal starting point for creating personas. Based on these personas, you will be able to create a more specific idea of the types of people that you are talking to with your chatbot.
Of course, a hugely important part of chatbots is, well, how they chat!
Use your user personas to gain an understanding of how you can programme your bot to talk in a way that is relatable and accessible to them. Is your bot likely to be used in a professional setting and is more formal language appropriate, for example?
On top of that, your user personas can give you insight into which industry your users are likely to work in, meaning that you can ensure your chatbot is familiar with any relevant jargon. In doing so, your user research will hopefully prevent a number of frustrating “sorry, I don’t understand” messages.
You can also continue to assess the types of language that is used in your chatbot later down the line once you have a mock-up or a prototype, before iterating your design based on your findings.
Empathising with your user
No chatbot is the same. Unfortunately, that means that there are no cut and dried formulas for how to create a good one that suits your users’ needs.
Take these two bots as examples: The “Genius” chatbot allows you to talk with “Einstein”, as a way of promoting a National Geographic television series. The Tesla chatbot allows you to check whether your car is unlocked from afar.
Those are two very different bots with potentially two very different users, right?
That’s why it’s so important to empathise with your user and know why they’ve come to your chatbot.
We may think we already know why our bot is being used but, often, we overlook important information and base our designs on assumptions. And that’s never going to end in a good user experience!
As Bianca Nieves highlights, understanding user needs for bots can be a case of utilising “any material you have such as customer service logs, frequently asked questions, online user review sites or comments sections on niche blogs.” to “Make note of common themes, unspoken assumptions, and end goals.”
In doing so, you will begin to empathise with your user, making you more equipped to create a chatbot experience that meets all of their needs.
Related article: Creating chatbots with great UX
Getting your chatbot into the real world
Whilst personas are powerful tools when you first create your bot, nothing beats getting out there and testing the product.
Get your users to complete a set of specific tasks on your chatbot and then ask them relevant (but not leading!) questions about their experience. In doing so, you can actually see how your users talk to the bot, any common problems that occur and, of course, anything that they enjoy about the experience.This can then shape future iterations of your chatbot to make them as good as possible.
Mockups are a great way to test your chatbot in the earlier stages. As Rucha Makati states, “They enable quick iterations with the user that allow us to validate the user experience and understand the user’s perception of the chatbot.” (Source) This is a key concept in lean design and within design thinking, as it allows for you to iterate, without spending huge amounts of money or wasting too much time.
Related video: Agile and Lean Startup for Beginners.
How chatbots could improve your user and market research
On top of needing user research to create good chatbots, chatbots could also help you to conduct good research! This is because of their ability for data collection.
Talking on data collection, Billy Leonard says that “By using a chatbot, you largely automate the process and — depending on how you incentivise participation — it can be mostly cost-free.
Additionally, chatbots allow you to collate data on the fly, and not have to wait months for your research to be analysed and fed back to you.” (source)
For example, Wizu is a chatbot surveys platform that is particularly good for getting customer experience feedback. It has options for creating NPS, CES and CSAT surveys and you can also create ad hoc surveys with things like sliders and multiple choice questions.
So, as you can probably imagine, it’s a brilliant tool for user researchers to have in their arsenal!
Data Sprints also summarise the potential for chatbots to be used within user research far beyond just surveys: “Chatbots also have interesting applications for shopper marketing, diary studies and experiential research. Instead of deploying a question and answer survey while a respondent is shopping or engaging in a behaviour, a chatbot could have a ‘conversation’ during the experience, recording all the thoughts and emotions as they happen” (source)
As with all digital products and services, user research is absolutely imperative in order to create chatbots that can reach their full potential. Not only will it allow you to create a simple, non-frustrating and maybe even (dare I say it?) fun, user experience, but it will save you wasted money and time in the long run.
On top of that, chatbots and user research are becoming more entwined and, as their AI capabilities only get more intelligent, we predict that they will become an even bigger part of user and market research.
The Snap Out Team ?