Why today’s chatbots aren’t more successful
Chatbots are all the rage in tech these days — Facebook is doubling down on bots in Messenger, and last year Microsoft went so far as to predict a coming gold rush for bot developers similar to the App boom of the late 2000’s. Messaging application Kik made them the central differentiator in the platform and received a valuation of $1 billion during their last round of funding. So why aren’t you using them?
Despite all the interest in the tech world, Chatbots have yet to fulfill on the “new app” promises with consumers. There is as of yet no gold rush for chatbot developers on Facebook Messenger similar to that which followed Apple’s app store in 2007, and most average consumers don’t even know what they are. Why is this? We’ve been watching chatbots fairly closely for the past few years and identified several problems in their current implementation that are keeping chatbots from really growing up.
First and foremost, many customers simply still don’t know that chatbots are available. Kik has done a good job of advertising them in their application, but Kik’s user base remains fairly small when compared to other messaging applications. Facebook likes to talk about how many people use messenger every day (and the numbers are truly staggering), but they aren’t doing much to make Facebook users aware that bots are now a big part of this platform or to help them find bots they might find interesting.
Even when customers are aware that a chatbot exists, finding them on any common messaging platform still proves to be extremely difficult. This is largely because most brands using chatbots already have some other presence on the platform that isn’t bot-related. Facebook is the worst offender — Messenger lets you find and send messages to any account on the platform, which includes brand pages that are often completely separate from the brand’s app. For example, Uniqlo has five different accounts on Facebook. Only one of these is a bot you can chat with — the rest represent regional brand pages and don’t readily respond to chat requests on Messenger.
To combat this companies are going to need a lot of messaging around where their chatbot can be reached. In the above example, every message sent to a Uniqlo Facebook account should have an automatic response directing the customer to try reaching out through the bot account instead (or the bot should just respond on all of them). Consider all of the ways you told your customers about the last new thing you made to provide value — reach out through those same methods with clear calls to action about where and how your bot can be reached.
Part of the reason most consumers don’t know about chatbots is that they have a fairly major platform problem. Microsoft CEO Satya Nadella said that ‘Chatbots are the new Apps’, so let’s consider that metaphor a little more deeply. The apps he’s referring to are the ones that exploded into the marketplace after Apple opened the iPhone to third party application developers in 2008, about a year after the iPhone itself was released. That first iPhone came with a number of embedded apps that provided a lot of obvious utility and made interacting with apps a natural behavior on that platform — the “phone” itself was an app, and presumably the one I was first and foremost buying the device for. Alongside the phone I had a camera, a calendar, and several other very obvious and focused applications that provided value out of the gate.
In contrast, bots aren’t first class citizens on most messaging applications — they share space or almost hide amongst my human contacts. In the phone metaphor, today’s chatbots are more like the bots you could reach on a brand’s 1–800 number than an app on a smartphone. Messenger, Skype, and even Kik aren’t futuristic smartphone-esque platforms for chatbot delivery — they’re old rotary phones designed to let you dial up your friends to see who wants to go hang out at the mall.
In the absence of a major messaging platform overhaul, companies are instead going to have to think about why a customer should go to all of the effort of reaching them on one of the existing ones. Most of the organizations with chatbots that we’ve investigated seem to have just assumed that their customers will look for them on messaging platforms. Deep investigation shows that most have given very little consideration to how the notion of “conversation” should change the brand experience, or the depth of value those conversations could bring.
For example, there are a plethora of personal shopping chatbots available, but almost all of them focus entirely on the shopping experience — helping the customer find what they’re looking for and facilitate a quick and easy purchase. That’s great, but it’s such a small part of the shopping process, and not the one most begging for a conversation with the brand. What about following up on how my items fit? What about letting me know when you have new items in my size in stock? What about getting feedback about an upcoming lineup?
Chatbots represent the ability to create a conversation, so companies should be looking more to the experiences given by their front-line sales associates or call centers to emulate rather than just making a sale. Your website probably makes sales just fine, and if not you have a different problem that a chatbot isn’t going to solve. Put another way, companies should stop thinking of “chatbots” as a separate channel and instead think about places to embed conversational experiences within existing channels of communication with customers.
As with any new platform for reaching customers, we can learn a lot of valuable lessons by looking to the past. Though the bot / app comparison obviously has some problems, the mistakes many brands made in rushing out apps with little value whose place in the customer journey wasn’t clearly defined should still be instructional for us; similarly the lessons we learned during the .com days when everyone rushed to create a fun, hip, trendy website.
Conversational interfaces with bots have a place in our future, but any foray into bot development today needs to be carefully considered. The conversations have be readily available, presented at times and in ways that make sense to the consumer, and provide value. You can best do this by following the same basic maxims you used to be successful on the web or in your native app: by carefully considering your customer, by doing the required research, by continuously iterating on the value you provide, and by measuring its impact. Without these things your bot will remain a cute little toy that few people ever find — a waste of time and money, and not worthy of further conversation.