This week I went to a talk and heard from one of the Alexa UK team from Amazon talk about their product in a way I had never thought about it before. I have an Alexa, and I enjoy talking with her and relying on her at times to help me out around the house, but I rarely thought about our interaction in the way it was described to me.

“Alexa hears what I say to her through one of a number of microphones, she transfers that information for it to be broken down, analysed and rebuilt so that an answer can be produced.” 

All of that is done in a fraction of a second and often with impressive and surprising accuracy.

The fact that I hadn’t thought about that as an Alexa user is part of why I recognise the quality of the usability and experience of conversing with her. And actually, why I should need to think about how clever she is when I talk to her.

But when implementing these technologies we have to look at it from the other side. Many technologies using AI almost assume these things, and as a result run the risk of creating chat and voice based interfaces that are truly impressive in their use of technology but lacking the sparkle that makes people want to interact with them.

Which brings me to the perfect example topic I find myself struggling with when I speak with Humans, the idea of what is meant by ‘this week’ and ‘next week’. But it’s not just Humans, when I had an AI assistant called Amy she couldn’t get her head around my use of these options at all. Ok, it is often used differently and varies depending on where you live, but what this demonstrates is to be truly able to support me, a system needs to understand not just what I am saying but the possibilities of and the variations in what I am saying.


Clients ask us: "How do we get our brand across in a Chatbot?

When we implement work with clients and talk about the effort required to create a chat interface, clients tell us that the topic or area being covered by the system is small and therefore must be easy to achieve as it only relies on minimal build. Then they say the fatal words – How do we get our brand across in this Chatbot? and the effort increases. We love these clients though because asking the question means that they get it and the end result will be far more successful.

The work needed to understand how and why we ask the question the make or break element in any chat or voice system and it comes down to the strong UX and Service Design starting point that new technologies need to be considered from.

Everything from the situation someone is in when they use the system through to having an answer for the inevitable questions of what is the capital of France, the system must be able to engage and steer people to a positive result.

At Aigen, we work with a focus on the technologies we build but we are always sure to do so with the user and the environment in mind. In fact, our Customer Experience background won’t let us forget it.

Do you want to find out more about Aigen? Say Hello!

Share this on: