Here’s What Chatbots Got Wrong

When Facebook launched enterprise chatbots in its Messenger app last year there was a ton of hype around the changing ways in which consumers would access digital information and services.

The technology was predicted to alter the face of entire categories from customer service to ecommerce to content discovery.

Unfortunately, now over a year later, much of this vision has failed to materialize. The reason is that the first generation of these offerings focused too much on style at the expense of substance.

Although there are over 100,000 developers who’ve built bots on Facebook, many of those early experiences broke down as soon as you scratched the surface of basic dialogue. As a result, retention rates proved underwhelming. Not only that, bots were difficult to find in the first place.

Discovery often depended on users learning about a chatbot outside of the social network and then migrating over in order to engage with it. The process was clunky and lacked the payoff to justify it.

Over the past few months, Facebook has moved to address these issues with two major adjustments to Messenger. The first is to integrate chatbots more deeply into existing conversations among friends (i.e., supporting existing discussions, rather than trying to start new ones).

The second is launching a discovery tab with recommendations for new chatbots to try. These initiatives are certainly a step in the right direction and will help overcome one of the biggest roadblocks to adoption — namely, that chatbots have been seen as marketing tools rather than utility-driven applications.

Large enterprises were particularly susceptible to this mistake. They watched the astronomical growth of platforms like Messenger and saw an opportunity to make their brands more “social” by joining in on the conversation.

Unfortunately, this just added to the clutter. “Standing out” on messaging platforms actually has less to do with visibility and far more to do with lowering the barriers to information.

In other words, success looks more like a personal assistant that already knows your preferences, rather than a talkative digital avatar.

Instead of trying to streamline communication between individual and enterprise, chatbots should have been trying to eliminate it in the first place.

As an example, if you ask Comcast what it’d prioritize when it comes to its social engagement strategy, the company would likely say “improving our dialogue with consumers.”

If you ask consumers what they’d prioritize, they’d likely say “never needing a dialogue with Comcast.” In this evolving dynamic between enterprise and customer, what matters most isn’t how talkative or friendly a brand can be, but how efficiently consumers can access its services.

Not surprisingly, much of the exciting innovation is being driven by startups, which are less encumbered by legacy approaches. For example, Trim will deploy a chatbot that can negotiate with Comcast on your behalf.

DoNotPay has successfully contested over 160,000 parking tickets across New York and London. And John Done will actually call local venues on your behalf to collect information via a synthesized voice. With a few keystrokes, we’ll soon be able to deploy a host of automated minions to handle more of our annoying daily chores. Cue ominous laughter…

In order to leverage messaging platforms as a growth driver, the next wave of companies will have to think more critically about the unique value they deliver on them and how to deliver that value as seamlessly as possible.

In conceiving new products, developers should first and foremost articulate why that product or service makes a user’s life easier.

One things is for certain — the old adage that customer is King [or Queen] remains truer than ever in digital. As more services become on-demand, power continues to shift away from companies and into the hands of consumers.

As a major driver of automation, bots may just be the catalyst that changes our discussion from B2C models to a new standard of C2B models.