The world has been overrun by bots, but that’s not really a bad thing.
Bots already account for more traffic on the internet than humans. An estimated two-thirds of tweeted links to popular websites are posted by automated accounts. And with the continuing controversy surrounding algorithmic influence manufacturing on social platforms, it might be hard to see anything positive about the increasing swarms of bots pervading the realm of human communications.
But Google’s recent demonstration of an incredibly human-sounding voice bot booking a hair appointment over the phone drew cheers from the audience, kudos from the press and has immediately sparked debates about how this technology will be applied, regulated and adopted.
The fact is that these often-conversational computerized incarnations of human helpers have very real merits that we’ve quickly grown to depend upon. They automate tiresome tasks and speed transactions so effectively that Juniper Research forecasts chat bots alone to be responsible for business cost savings of more than $8 billion per year by 2022.
And it’s not just businesses embracing bot benefits. One recent report found that 45 percent of end-users prefer chat bots as the primary mode of communication for customer-service inquires. Other statistics show that the vast majority of consumers who have interacted with bots found them effective for resolving customer-service issues.
And demographically speaking, bot interaction is practically second nature for younger generations. Research shows that 70 percent of millennial consumers report significantly higher preferences for transacting via mobile messaging (including messaging applications, texting, social media and chat bots), and that millennials will use bots twice as often as consumers aged 35 through 54 and six times more often than consumers aged 55 and over.
More and more people are engaging with brands through bots every day. At Facebook’s F8 developer conference in early May, the company reported 300,000 active bots engaging with customers on the platform’s Messenger service, and more than 8 billion messages exchanged between people and businesses on that app monthly—a fourfold increase over last year’s numbers.
If you follow technology trends, you may start hearing a lot more about “omnibots”—comprehensive artificial-intelligence-fueled digital representations of brand identities. These computerized personas transcend mere customer-service or social-media chat bots but are more domain-specific than the consumer voice-enabled digital assistants that have become so popular in recent years.
Think of an omnibot as a conversational natural language processing system similar to Apple’s Siri or Amazon’s Alexa, but also text and/or video and/or virtual reality-enabled—and focused specifically to front a soda company or a bank or a clothing store. These omnibot brand embodiments are being designed as a ubiquitous public interface integrated into advertising, apps, websites, social media, call centers, stores—you name it.
Bots are here to stay.
When they work as intended, they are marvels of modernity. Starbucks’ barista bot lets you chat custom orders and make payments from your phone. Kayak’s bot will tell you if there are available hotel rooms in London this weekend and what a last-minute flight will cost. Capital One’s Eno bot lets you check account balances and pay your bill via chat. The Western Union bot lets you send money around the world from your phone. And the Whole Foods recipe bot will suggest what you should cook for dinner tonight.
And as bots increasingly serve as the point of contact between brands and customers, they are becoming more than mere high-tech cost-savings and convenience enhancers. They are becoming the public face of their brands.
And therein lies the danger.
When bots don’t work as intended, brand reputation suffers. The internet is littered with stories of bots that are poorly implemented, frustrating, pushy, context-impaired and creepy. “Sorry, I didn’t understand” is not a catchphrase you want associated with your business.
Moreover, when bots are responsible not just for novel informational services, but for representing the sum total of the customer experience, your brand is on the hook for misunderstandings.
These catalogs of negative interaction mean that customer experience suffers when bots fail. And as we all know, CX is a crucial differentiator for brands, but that isn’t the only hazard.
But what if Alexa started providing incorrect information or broadcasting sensitive information such as credit-card numbers or passwords? To my knowledge, that hasn’t ever happened, but could it? What if a national sports team’s social media bot started sharing racist content (that has already happened). And what if an automated financial system program “goes rogue” and starts executing unauthorized transactions (that already happened, as well).
Human representatives make mistakes that reflect poorly on their brands, as well, but when ubiquitous automated systems make those same mistakes, the impact is amplified exponentially.
To mitigate such reputational risks, brands must apply the same safeguards to bot function and behavior as they do to managing traditional representations and representatives of the company:
Start with knowing your customer: Understand their needs and expectations clearly and design any automation solutions specifically to perform those functions. Real utility is the hallmark of any great bot.
Incorporate journey mapping: Take the time to create a visual design of every customer touch point with your business and where any bot interaction will transpire. Start at the beginning—how the customer is acquired—and create a map of each step taken with the product or service. View your processes from the customer point of view to understand how bots will be perceived.
Use the right technology: Be certain that you have the ability to supply all of the pieces required for a successful bot deployment. Do all of your engagement channels function as a single system that shares interaction data and insights, and can bot technology be integrated? Can the bot seamlessly transition to a human for more complex support? Are there safeguards to vet customer feedback, complaints and bot engagement data and a system in place to manage failure?
Additionally, your audience/customer should be informed when they are dealing with a bot. You want to set expectations and reduce the possibility of confusion if the human thinks they’re able to converse openly, but the bot is only able to perform a limited set of functionalities.
As we move into an increasingly automated future, our new technological capabilities offer a lot of promise for productivity and convenience and engagement. But not everything can or should be “botified.” Brand responsibility cannot be automated, nor can reputation. If your bots fail, so do you.
Chris Connolly is vice president of product marketing at customer experience platform Genesys.