There has been a lot of hype around chatbots since they burst onto the scene several years ago. The idea of interacting with technology through natural human language once seemed inherently futuristic and exciting, perhaps even a little gimmicky.
Today, chatbots are just another technology that we use. It has become normal for our everyday lives to be augmented by smart assistants such as Siri or Alexa, and we’re just as likely to meet a marketing chatbot as a real human when interacting with brands online – indeed many now prefer it this way.
So now that the dust has settled and the ‘wow’ factor has subsided, what can we make of the chatbot revolution? Are chatbots a passing fad, or are they here to stay?
Beyond the hype
Countless industry experts called 2018 the year of the chatbot, but then they said the same about 2017, and 2016. For the last few years, chatbots have been the subject of much hype, as expectations peaked about their potential to revolutionise our lives.
Now that it’s 2019, it’s safe to say that the year of the chatbot, whenever that was, has come and gone – but chatbots are here to stay.
To better understand where chatbots are on their journey of development, let’s take a look at the Gartner hype cycle – a graphical visualisation of the typical path a new technology takes, from conception, through early adoption, to eventual mainstream use.
As you can see, chatbots have been through the initial peak of inflated expectations, where people get carried away on a wave of hype, and through the subsequent disillusionment, when reality inevitably falls short.
We are now at a stage where expectations are more aligned with reality, where we understand both the benefits and drawbacks that chatbots bring, making it much more like a tool in our back pockets than a revolution.
As we pass through the initial hype stage and out the other side, we gain a better, more realistic understanding of what chatbots are good at, and what they are not so good at. We understand their limitations, which helps us to maximise their performance of tasks at which they genuinely excel.
Like any tool, chatbots are naturally designed to do somethings but not others. The key to making the most of chatbots going forward it to understand where that line between success and failure lies.
So what are chatbots good at?
Chatbots excel in transactional exchanges, where they make once complicated processes as frictionless as possible. That means less time spent, less hassle, and fewer steps involved. This could be anything from helping a customer to book a flight online, to providing a single interactive platform between employees and a number of technological systems.
Without a chatbot, booking a flight is a relatively lengthy and complicated process that involves navigating websites, clicking through pages, and filling in forms. With a chatbot, however, these processes can be completed within the space of a few short sentences. All you have to do is text or speak, and the chatbot handles the rest.
Chatbots are not only good at receiving and processing instructions; they are also able to push out information to the user in a more natural way. By gathering relevant data and finding patterns, chatbots can provide useful insights through automatic push notifications. This way, chatbots can actually advise the user in an active way, rather than simply responding in a passive way.
And what are they not so good at?
Given that chatbots generally simplify processes down to a short conversational exchange, they are not ideal for presenting large amounts of information, multiple images, or complex dashboards. In these cases, a more traditional approach to web navigation is likely to result in better user experience.
While chatbots are excellent at gathering information, there are times when traditional web-based forms are still a better option. The conversational nature of the user experience makes short interactions ideal, but if a process requires a large amount of information to be gathered, this back-and-forth exchange can quickly become tiresome. In these cases, users may find online forms easier.
What about the future: have we reached peak chatbot?
The initial hype phase may have passed, but don’t be fooled – the influence of chatbots will continue to grow. Most of us are now accustomed to the idea of talking to a bot, but in the future, this process will play an even more central role in our daily lives. Astonishingly, Gartner predicts that by 2020, the average person will have more interactions with a bot than with their spouse!
It’s important to remember that the technology that drives chatbots is still under development. As improvements in natural language processing (NLP, i.e. the ability of a bot to ‘understand’ what we are saying) and natural language generation (NLG, i.e. the ability of a bot to turn data into language that we understand) improve, the scope of chatbot usage will grow. As with the internet and smartphones, perhaps we’ll look back in a few years and wonder how we ever survived without them.