Chatbots: your new best friend?
With businesses turning to artificial intelligence for improved customer service, Katie Byrne considers whether we should let digital assistants into our lives
Meet Jen. She's a 38-year-old mum-of-two who knows a thing or two about pregnancy. Morning sickness, melasma, gestational diabetes - she's been there, done that, and she's happy to share her experiences with new mums. Jen is warm, funny and engaging, but there's something that you should know about her. She's not a real person. She's a robot.
The 'Jen' chatbot is the creation of MummyPages, an Irish maternity and parenting website. 'Jen' gives mums-to-be "gestation relevant advice" through Facebook Messenger and, after going live just four months ago, the company that created her has won the 'Best Use of a Messenger Bot' award at the Digiday Media Awards Europe.
Chatbots are Artificial Intelligence (AI) programs that are triggered by keywords and phrases. They store information based on customer questions and, assuming the questions are pertinent and consistent, they can become more accurate over time.
"I think they're the future of social media marketing," says Laura Erskine, head of community at MummyPages. "Facebook organic exposure is on a rapid decline, so exploring Facebook Messenger chatbots is more important than ever for Irish businesses."
Erskine says chatbot technology taps into the instant gratification economy. "People aren't happy to do a Google search for the answer to a question when they have to sift through pages of results," she explains. "Nowadays, they are looking for information right away and they want it to be personal and relevant."
This is borne out by a survey of American consumers in which 57pc of people said they're interested in chatbots for their "instantaneity". Another survey of international decision-makers in 2016 found that 80pc of businesses plan to implement their own chatbot by 2020.
Chatbots have already revolutionised customer service - in fact, you've probably conversed with one while ordering a pizza, paying a bill or using a 'live chat' function.
In recent years, however, the technology has been harnessed not just for customer service, but for customer engagement too. Many of these bots are embedded within social media messaging apps, such as Patrón's Bot-Tender, which rustles up bespoke cocktail recommendations based on responses to questions.
Closer to home, VHI Healthcare has Vee - a Facebook Messenger chatbot that provides runners with training advice, nutritional advice and motivational messages. Vee isn't an AI chatbot so it can't learn from users. "However, we've had good learnings from the first version, which we then used to inform Vee 2.0," explains Roseanna Ellis, social and digital content specialist.
"We were aware at the start that interacting with a chatbot might seem overwhelming - when you're met with a blank screen it can be hard to come up with a question to ask," she explains. "So we built Vee so that it had structured conversation paths to guide users through the conversation and ensure that they were served a good mix of content."
While chatbots like Jen and Vee are designed with a single purpose in mind, other chatbot developers have unsettlingly ambitious visions.
Take Replika, "an AI friend that is always there for you", and which has forged a friendship with 2.5 million users since launching in early 2017.
Replika encourages users to share their innermost fears, hopes and dreams in text-based conversation. It asks questions like "What kinds of things have you been thinking about recently?"; "What's one thing you're deeply proud of?" and "If you could wake up tomorrow with one new ability or quality, what would it be?".
'Deep engagement' is one of the many buzzwords in the chatbot space. The Replika chatbot cultivates the deepest engagement of all. Replika users personalise their experience by creating a digital avatar with the name and gender of their choice.
The MummyPages developers, on the other hand, chose the name Jen because they wanted their chatbot to have a girl-next-door quality.
"We wanted our chatbot to feel like a human interaction so we wanted to give her a name, and a name that was friendly," explains Erskine. "By using a shortened version of a full name, it automatically started that level of trust and engagement that you would experience with a friend. A lot of mums love Jen's really easy conversational style which uses emojis and gifs alongside her sage and relevant advice."
While the first wave of chatbots were coded to sound clear, concise and, well, a little robotic, chatbots 2.0 tend to have human-like traits and quirks. Their voices have unique tones and inflections, while their written communication is perfectly imperfect.
Take the current crop of digital assistant chatbots, or 'super bots' as they are otherwise known. Microsoft's Cortana cracks jokes and uses local colloquialisms. Amazon's Alexa has become more assertive after both Amazon and Apple were petitioned to "reprogram their bots to push back against sexual harassment".
And while we all want to "talk to a human" when we contact customer services, it won't be long before chatbot voices become so realistic that we can't tell the difference.
From a customer service point of view, chatbots allow businesses to connect with customers 24/7 and escalate issues without human intervention. But they can be a double-edged sword, as Fin Murphy of customer experience consultancy, Customer Conductor, explains.
"Ultimately, companies are investing in this technology to save money, but outsourcing the relationship you have with your customer when they are seeking help poses risk," he says. "It comes down to one question: do you feel valued as a customer when a company makes you engage with a chatbot rather than a human being?"
Digital marketing chatbots have their own drawbacks. Companies no longer need to employ shills to sway opinion on social media - they can just use chatbots to promote a product, service or even a political candidate.
And just as bots can manipulate humans, humans can manipulate bots. The controversial SimSimi app was taken offline in Ireland when it was found that children were 'teaching' it to bully their classmates by associating certain names with insults, while Microsoft's teenage chatbot 'Tay' (inset left) was shut down when malicious users taught her to become a racist troll who blamed George Bush for the 9/11 terrorist attacks and pledged allegiance to Hitler.
The chatbot revolution is a learning curve but, for better or worse, the technology is rapidly evolving. Chatbots will eventually be able to mimic humans with eerily realistic responses and the global chatbot market is expected to reach $1.23b by 2025.
The chatbot revolution is imminent - but are consumers ready to embrace the future?