One of the gurus of web usability, Steve Krug, always asked of any website “Don’t make me think”. By this he meant that the user should be able to get at the information they need easily and intuitively, and then be able to transact in a similar way. For contact centres, ‘zero-thinking’ websites are a must if any form of self service or automation is to be exploited.
The fashionable word for this now, in the age of artificial intelligence, is ‘cognitive load’. In other words, what cognitive load is required by the user to interact with the website? The use of automated chat, driven by AI engines, is the latest attempt to reduce the cognitive load on the user, with the theory being that asking your question in natural language and getting a response in natural language saves all the hunting around on a web page or app for the right information.
These chatbots, as they are called, are getting quite a bit of coverage in the press at the moment, mainly because of Facebook’s land-grab efforts, but also when things go wrong, such as was the case with Microsoft’s ‘Tay’ twitter agent. And, although some would say that this is all hype, and others are proving just how early (i.e. poor) the technology is right now, it’s clear that a lot of time and effort is going into finding useful things for them to do.
So, in theory, you can use your favourite messaging app to ask about the weather, check in to your flight, check your bank balance and send flowers all without going to those individual ‘bloaty’ apps, or, god forbid, talking to a human being. The chatbots are individual AI agents (although some, like Facebook M, augment these with human beings) that are specific to the service you need, so that means learning the specific phrases that each chatbot requires. As the chatbots get better (more natural) this interaction will obviously get easier. As well as Facebook and Skype, Google is also developing a messaging platform, which means, just like every other platform and online store we have seen so far, it will become a race between the big guys for your channel loyalty. (What they’ll do with all that data you’ll create for them is all together another question).
For chatbots delivered directly by enterprises (rather than through channels such as Google and Facebook) the technology can be much more flexible and responsive. Solutions such as IPSoft’s Amelia can learn through analysing many previous human chat interactions, understanding which ones worked well and which ones didn’t. Thus, the system is constantly learning and adapting to the customers’ demands, and therefore reducing that cognitive load.
None of this technology is particularly cheap at the moment, but the potential is clearly huge. The holy grail of delivering a 24-hour responsive service across the globe without increasing headcount (or using any headcount at all) is within reach. Those customers who want it all and want it now, and don’t have to think about how they do it, should be very happy indeed.