If you’re not familiar with chatbots already, the clue’s in the name. Chatbots, or conversational agents (or even dialog systems), are programs that mimic human interaction, encouraging visitors to a site to hold online conversations with them over a messenger service. They are fast becoming a crucial cog in businesses’ online marketing and customer service machines. In fact, following the industrial revolution’s disruption of manufacturing, the advent of consumer internet and then the dawn of social media and mobile phones, many hail conversational robots as the next obviously revolutionary tool to have a serious impact on business practices. Hence why big tech companies like Microsoft Google, and Facebook are making huge bets on chatbots.
Chances are, then, that you’ve had a brush with a chatbot yourself. If you’ve bought anything online or submitted a minor customer service complaint, it’s highly likely that a chatbot was on the other end of the interaction. Chatbots are partly valuable for their standardised tone – sulky customer service representatives with poor language skills and a tendency to audibly lose their temper no longer need to enrage consumers, because chatbots are on hand to deal out consistently charming platitudes. That’s not to say there isn’t room for improvement – glitches in the technology can often lead to misunderstandings, nonsensical answers, and a failure to execute the task in hand.
But the chatbot appeal comes largely from the fact that, as we become increasingly accustomed to using messaging apps for communicating with everyone about anything, the chatbot presents an opportunity to engage in dialogue (without waiting) with a ‘representative’ of the business. The implication is that our questions will be answered sooner and our problems will be solved faster. To the businesses’ great benefit, chatbots can be programmed to hold conversations simultaneously with numerous (and sometimes limitless) members or customers. Once the initial investment is done, the maintenance costs on chatbots are far thriftier than the restrictive activity of an expensive and potentially unreliable human-only support team.
Chatbots also match the quirks of a new generation – the “millennials” for lack of a better term – and their deep-seated fear of making or receiving calls. An instant messaging-based conversation with a robot is a far more pleasing idea than a phone call with a stranger when 90% of your social interactions have been guided by a thumb swiping across a keyboard.
In addition, chatbots enable the seamless use of multiple apps from one location. In these instances, there’s no need to close the text or voice conversation to follow a process like this:
Pull up a map to a restaurant > look at the menu > make a reservation.
The entire series of steps can be carried out from a single interface, making the experience quick and efficient as well as keeping the user engaged – unnecessary breaks in the experience or distracting changeovers are likely to increase drop off rate.
Current productivity uses for chatbots can see them put to work as internal AI assistants, scheduling appointments and follow ups, filing expense reports and carrying out other basic administrative tasks. Large businesses are also finding use for chatbots when it comes to monitoring sick leave taken by employees, or keeping tabs on individual and team workloads, while presenting a user-friendly voice at the other end.
But at this precise moment in time, chatbot technology is only really likely to revolutionise your business efficiency if you’re engaged in a considerable amount of customer service activity. Facebook, for example, is allowing businesses to interact with their consumers via the Messenger app. Banks are using chatbots to deal with fairly straightforward customer service issues like lost and stolen cards, and highly developed chatbots are available in the familiar shape of Virtual Assistants like Apple’s Siri, Microsoft’s Cortana and Amazon Echo’s Alexa. And with the Internet of Things projected to include over 50 billion connected devices by 2020, chatbots will contribute to the functions of smart cars, voice navigation on wearables and other internet-enabled devices.
If the frontline interaction and customer service benefits of chatbots suit your business best, you have a range of choice in terms of type and positioning:
- Query bots can answer questions about your services, opening hours, discounts and offers
- Shopping assistant bots can help consumers identify the item they’re looking for, or take them on a verbal tour of your business’s e-commerce stock
- Explainer bots can give users a richer and more detailed experience of online shopping by giving more information about how a product works and what it’s like to use it
- Post-purchase bots can deal with complaints, returns, exchanges and feedback.
Despite their benign moniker, chatbots have their dangerous side, too. Last year, Microsoft launched an experimental chatbot called ‘Tay’ on Twitter, which was programmed to deal with a high volume of communications. Tay’s artificial intelligence learning was dependent on the material it consumed online, so trolls were easily able to train Tay to make offensive remarks, and it was hurriedly taken down. These gaffes are potentially much more damaging than an irritable and overtired call centre operator but more importantly just embody the philosophy of “garbage in, garbage out” and how much freedom AI should be allowed to develop, especially given Stephen Hawking’s and Elon Musk’s concern about the ‘Singularity’.
Difficulties lie ahead in relation to privacy, too. The more customers share with chatbots, the more data the business has access to. Complicated debates surrounding data security and privacy rights will inevitably arise, and will prove unsurprisingly difficult to legislate.
The ultimate goal however is to develop chatbots that can have natural conversations indistinguishable from genuine human interaction. Many experts are putting their faith in Natural Language Processing and Deep Learning techniques, although given the hype around AI is can be difficult to separate fact from fiction. Despite this, with many computer scientists journeying towards human-level machine intelligence aboard the S.S. Moore’s Law we should accept the rise of chatbots as a case of when, not if.
It’s likely that plenty of time and money will invested in developing the technology but the chatbot’s place in the digital business technology hall of fame is dependent on significant developments, and without meaningful progress they may drift out of relevancy.
For the time being, though, businesses seem keen to maintain loving relationships with their new partners in crime, and will continue to invest as the opportunities multiply.