It was revealed last year that a staggering 44% of US consumers would now prefer to communicate with a chatbot over a human for customer relations purposes.
The findings will reassure many businesses which placed big bets on chatbot tech over the last year or so, and may convince laggards to take the plunge, too: reportedly 80% of businesses expect to be chatbot-enabled by 2020.
The potential of the chatbot is so great that businesses are deploying the technology to handle customer complaints, take orders, recommend products or services, and more besides.
So could 2017 be ‘the year of the chatbot’?
Chat app usage for the likes of WhatsApp, Slack and Facebook Chat has now eclipsed that of social networks. With massive audiences on these platforms businesses can use chatbots to bring their product or service to bear on a new format which their customers are now naturally familiar with.
However, doing so successfully will require businesses to navigate a few tough challenges, such as:
Do you really need one?
Amid the hype, it can be hard to decipher just what a chatbot can do right now, and what it can’t.
Chatbots work from rules and process – while some can only respond to specific commands, others are more flexible, using machine learning technologies to adapt their pool of responses to new language.
Those leaps in more nuanced communication make for great customer service roles. Some have a capability for sentiment analysis to respond appropriately to angry users, and some use natural language processing techniques to give more natural responses.
However, at this stage there remain natural limitations to a bot’s intelligence, so early adopters should keep a tight remit on their tools. If your bot can take shoe orders, for instance, but can’t yet handle customer complaints, don’t name it an ‘assistant’ as that could imply too much about its abilities.
Managing user expectations at the outset can save a lot of confusion and frustration further down the road.
Fear not the blundering bot
Despite widespread interest, some earlier iterations of chatbots have given sceptics good reasons not to adopt the tech. There is a perception that automation is a source of danger for brands – overcoming that will be essential to reaching the 80% adoption figure quoted above.
But are mistakes really such a great threat to brand reputation? According to Dylan Stuart, partner at global creative consultancy, Lippincott: no.
Dylan’s take is that chatbots might work well for brands willing to cede control on the specifics in favour of building an emotional connection with audiences. It might entail some faux pas but, after all, to err is to be human.
To chat, snap or something else?
Now that brands are well accustomed to image-first social platforms like Snapchat, Pinterest and Instagram, some innovative businesses may well be considering how to use the immersion and immediacy of those platforms with chatbot tech.
Images can be processed faster by the brain faster than text – 13 milliseconds, according to MIT. So might they play a big part in urgent communications, or in making complex conversations straightforward for the user and bot alike?
Ed Preedy of GumGum, a computer vision platform for marketers, recently discussed the possibility of computers and users conversing with images by enabling bots with image recognition abilities. Just as businesses were getting used to text bots, a whole new world of opportunity may be around the corner.
Will chatbots usher in a new human-computer language by the end of the year, or will we still be struggling with bots gone rogue? We’ll be following the ascent of chatbots closely in the months ahead.
@SimpaticoPR – Intelligent Business PR