Facebook's Messenger bots: it's tough to make AI seem human
Facebook's Messenger bots: it's tough to make AI seem human
A view from Ali Strayton

Brand bots on Facebook will need empathy and humanity to work

The robots are well and truly on the march, with their herald Mark Zuckerberg (and pretty much every digital media outlet on the planet) proclaiming...

As a consumer, I’ll appreciate a bot that quickly gives me the information I’m looking for, or a deal at a restaurant where I’m meeting friends after work

To those of us with even a passing acquaintance with social media, chatbots are hardly a new phenomenon. Twitter has been infested with them for years, and they are becoming an increasingly difficult kill – as this recent research found.

Of course, there’s a difference with the bots of the new ‘b-commerce’ economy: they are opt-in, giving them a huge advantage over both unwanted social bots and the apps they are supposed to be replacing. The challenge for app makers was always two-fold: get the user to download the app, then get them to use it more than once. A bot integrated with an existing platform neatly sidesteps that problem by becoming a seamless part of the user’s experience.

And once nestled snuggly inside the user’s always-on digital ecosystem, a bot brings a potentially potent level of brand interaction. Witness the success Disney had with its Miss Piggy Facebook chatbot. Yes, it was pretty primitive but it certainly shows the direction of travel, and the more data the designers have, the better they should become (Microsoft’s recent racist chatbot problems notwithstanding).

And they’ll certainly have to get better, because to today’s sophisticated consumer, bots will be seen as yet another transaction in the great data trade-off – I give you information about me, you give me something that makes my life better. As a consumer, I’ll appreciate a bot that quickly gives me the information I’m looking for, or a deal at a restaurant where I’m meeting friends after work.

Lacking the human touch

But utility is one thing, brand empathy is another and here is where things get a little more complicated for marketers. Even the best-scripted bot lacks the sophistication needed to emulate genuine human communication; indeed, the closer they get, the further down the dangerous uncanny valley they travel. They simply don’t understand the nuance, or humour, that is central to our social interactions.

Human-to-human connections will remain the best way to generate genuine empathy for a brand by, for example, making people laugh. It’s like asking whether the best way to understand comedy is from repeated viewings of a Michael MacIntyre DVD or a live after-hours performance in a sweaty pub cellar.

A bot’s pre-programmed, committee-scripted comedy stylings might, at best, be able to elicit a chuckle or two on first encounter – and offer absolutely no value on repeat viewings. A live stand-up’s unscripted put-downs of drunken hecklers, however, will generate the more visceral emotion – one far more likely to make a long-lasting impression.

The real power of social media to build relationships comes from responsive interaction. A properly handled response to a customer interaction not only has the potential to win over an individual, it can turn them into a genuine advocate. Ultimately, that is of more value than a flurry of automated purchase suggestions or discount vouchers.

What the bots will do for smart brands is act as both a filter (keeping low-level customer response work away from your staff through automation) and a signpost (pushing users towards content you have created). This is where they will really add value, by freeing up your flesh-and-blood marketing or customer relations teams to deploy their greatest asset – their humanity.