Many have already declared that 2017 will be the year of the chatbots. Now, we're not incredibly far into 2017 yet, but the chatbot buzz is still strong and quite a few things have changed.
Early on, as one of the biggest chatbot proponents, Facebook launched their Messenger platform with chatbots and expectations of rapid adoption.
While they were right in one aspect -- Messenger has quickly grown to have the largest user base of any social media platform or chat app -- their chatbot hopes and predictions might be a little ways off.
The Year of The Chatbots
Why, then, is everyone praising 2017 as the year of chatbots? As many have already pointed out, even Gartner has begun pumping out content surrounding bots, "Four Use Cases for Chatbots in the Enterprise Now." Well, the truth is, chatbots might not be doing as poorly as some thought.
In reality, chatbots are "entering the market rapidly," according to Gartner. Chatbots have a wide appeal for both everyday consumers and users, but also the businesses that include bots into their mobile app strategies.
Gartner clearly thinks bots are here to stay, even now that we're almost a quarter into 2017. So with Gartner pushing content, and Facebook recently announcing even more Messenger and bot changes, we wanted to take the time to check up on the state of chatbots in 2017.
A Change in Semantics
To gain a stronger understanding of how well bots are doing now, we first need to dive into the origins of these friendly, faceless, and helpful companions. But first, I want to echo a distinction made by Facebook Messenger VP David Marcus at Facebook's F8 conference this year:
"We never called them chatbots," Marcus said, "we called them bots. People took it too literally in the first three months that the future is going to be conversational."
This is an important distinction that will help frame our understanding of progress and the changes made, as well as some of the disappointments.
Where It all BeganJust about a year ago, Facebook came out swinging with its Messenger platform, and what was referred to as "bots on Messenger." The provider said that Messenger combined with bots could actually boost loyalty and customer engagement, and it launched with a pretty extensive list of chatbot partnerships. Certain developers gained early access to the platform; most notably at the time was 1-800-Flowers.
This, of course, was all made possible thanks to AI and natural language processing. Zuckerberg had even said that AI combined with some human help will allow users to hold fairly normal conversations with bots just like one would with friends. So, this comment is most likely why the "conversational" aspect took off and was a major outside focus.
The announcement went on to explain how the Messenger Platform's Send/Receive API allow bots to send not only text, but structured messages. In fact, this is the exact idea that RCS brings to the table -- more advanced messaging. But for Messenger, it's built in to begin with, which explains some of the power behind the platform. For discovery, there was a search bar for bots, and companies could easily shift SMS messaging users directly to the Messenger platform. Bots combined with Artificial Intelligence are revolutionizing our customer experience.
Bots Are Here, But Facebook is Big
Facebook introducing bots wasn't some new, revolutionary tech. However, it was a massive mainstream player making a massive move.
Overall, the sentiment was that bots are here, the technology is impressive, and this should help us revolutionize the way users interact with agents and customer service as a whole. But the main point -- the main selling factor of why Facebook is so important with bots -- is the sheer volume of Messenger users.
In fact, about midyear 2016 it was reported that Facebook Messenger has grown to a massive 1 billion users. This is without a doubt one of the largest gatherings, or followings, in all of history. Maybe the Catholic church compares, but this is really staggering. And this is exactly why Facebook is the highlight of chatbots. Everyone uses Facebook, and everyone uses Messenger, so the thought is then that everyone already on these platforms will also use bots.
But this is where things get a bit more confusing, with Facebook making some interesting changes recently. Most notably is one I reported on when announced: Messenger added in-chat menus for users to navigate bots without conversation. And here we come full circle to the more recent, "Bots not Chatbots," comment.
The Gartner Hype Cycle
I want to start this off by addressing that I'm not necessarily the first to make the connection and identify chatbots as falling into the Gartner Hype Cycle. A great series put together by Paul Boutin for Chatbot Magazine specifically highlighted the Hype Cycle and its relevance to bots. Boutin wrote:
"Last year, chatbots hit a well-known apex of irrational exuberance. Technology watchers refer to the Gartner Hype Cycle, a roller-coaster journey from new technology to everyday useful tool. Gartner, the company that canonized the idea, is one of the leading industry analysis and forecasting firms. Over the years, Gartner analysts have seen one new tech after another follow this path"
In fact, this phenomenon is so well recognized that Gartner themselves have their own explanation, and drew up this nifty little illustration. Just last year, Boutin argues, chatbots were at their Peak of Inflated Expectations. This explains all of the, "2017 will be the year of chatbot," discussions. And just because chatbots are following a normal hype trend doesn’t mean these predictions are wrong.
But I like this point a lot -- I think hype is a massive factor in a lot of what we hear about. I enjoy Slack, but I'm no stranger at pointing out that they rode a pretty massive hype train to success. Now, credit where it's due, Slack came out at the right time with the right combination of fun and productivity. Innovation isn't necessarily a new idea -- just the right combination and timing. Whatever works, right? Except for when the hype leads to disappointment.
So Facebook and the Messenger platform served as the Innovation Trigger; media hype raised the concept to the Peak of Inflated Expectations; and more recent Facebook changes as well as media assessment has put us back down into the "trough of disillusionment." But we're not far from "Slope of Enlightenment," and working our way towards productivity.
Just look back at Gartner's 2016 Hype Cycle from July. We can see "Conversational User Interfaces," "Virtual Personal Assistants," on their rise UP to the Peak, and more importantly, "Natural-Language Question Answering" already sliding its way down to the "Trough of Disillusionment."
So, why this change in hype, and what does this mean for chatbots?
Where Are We Now
Well, first we need to understand where we are, and what has changed since major bot adoption. I think Facebook's keynote at F8 2017, an exact year after the introduction of Messenger and bots, is a great focal point for this discussion.
One of the biggest statistics to recognize here is that Facebook themselves report that, in just a year, Messenger has grown to support 100k developers who have made 100k bots. Now these are probably nice round numbers to keep everything neat for presentation, but this is also a big jump over the 30,000 numbers we were hearing in April of 2016.
Also, prominently displayed right on Messenger's homepage is the user base of Messenger, which has grown to 1.2 Billion (as if it could get any bigger). Facebook also highlights 2 billion messages are sent each month, and "53% of users say they are more likely to do business with a business that can message."
But Numbers Don't Tell All
If you ask Facebook, they would probably tell you that bots are doing very well. But moving away from just raw statistics, Marcus also highlighted some specific organizations that have seen a benefit from bots during his presentation. According to Marcus, Canada's Rogers "reported a 60 per cent improvement in customer satisfaction following the deployment of a Messenger customer service bot."
Or another quote from the same coverage that I felt was equally important:
"Marcus also mentioned that Sephora bought ads on Facebook's Newsfeed that directed people to Messenger, resulting in an 11 per cent increase in makeup booking sessions."
This illustrates how chatbots combined with traditional advertising can help boost adoption. I think the key takeaway here is that people will follow where their favorite businesses go. The idea is to go where the people already are, and everything should fall into adoption naturally.
A Shift in Bots: From Text to GUI and Back Again
But what about all those recent changes? Why did Facebook introduce menus and stray away from the conversational aspect of it all? Well, Facebook addressed this both at F8 this year and in a blog post leading up to the conference.
Without going into too much technical detail, Facebook makes a distinction between text-based and graphic-based user interfaces. The old style command line, text-based operating systems worked for very specific functions.
Meanwhile, GUI (visual-based) operating systems allowed for even more control when conversation just wouldn't cut it. As Google explains, if you wanted to zoom in on a GUI, you simply drag your mouse or pinch your fingers. But with a text OS, you would have to type in a very specific "zoom 300x" command.
So the key is that neither is better, just better for different things. Facebook wants to marry the two methods of interaction directly in their bots -- again, why they don't want to call them conversational only. Not everything can be done through conversation, so the menu is there to help fill the gaps.
Gartner seems to agree as well, as shown in their list of ABC Technologies that will change the future of the customer experience:
"This technology offers a new human interface, which is particularly interesting since much of the current human interaction with technology is through a screen, a very old-fashioned approach. Interacting with chatbots won’t require any particular setup; the technology will simply understand and do as the human asks."
Do Different Numbers Tell a Different Story?
Now of course Facebook is going to promote their own platform and their own solutions. They wouldn't share numbers if they weren't very promising. And that's not to say other numbers contradict Facebook's open mind to bots. As I noted, Gartner still feels bots are a relevant force and will continue to remain one, otherwise they wouldn't even put together a report on how to use bots.
But during my research I came across an interesting report. The 2016 Mobile Messaging Report was put together by Ubisend, a mobile messaging platform provider. There were a few statistics in here that I thought were relevant to the chatbot discussion.
My main takeaways focus on just a few statistics:
- "Over 50% of people surveyed said it is important for a business to be available, contactable and responding promptly 24 hours a day, 7 days a week."
- SMS is more appealing than phone support to one third of people surveyed, but website online chat is even more preferred over SMS. In fact, "80% surveyed said they would utilize online chat over SMS."
- "49% of the people surveyed said they would rather use a messaging application to communicate with a business than a phone call. Over 60% of consumers believe businesses should be available, contactable and responsive through messaging apps."
The reason I like these numbers is because they highlight directly what consumers are looking for. After all, consumers are the ones we want to get using our bots. Basically, most people prefer to chat online, or even text, instead of talk on the phone. But as messaging apps became more and more popular like Facebook Messenger, users would prefer to stay where they are. Again, this is why Facebook is such a key player here.
So Where Does This All Leave Us?
Well, first thing is that I think most predictions were correct. Chatbots, or bots in general, are here and they are here to stay. On its own, Artificial Intelligence is already a powerful force behind some really in-depth big data analytics. Combine the two emerging technologies of bots and AI, and we can have a really robust recipe. If you don't want to take my words for it, here's what Gartner had to say:
"A bot is a small piece of code or an app that does one thing really well. It’s possible to create bigger, more complex apps by combining multiple bots and building them up as you would with lines of code. AI and bots give technology a voice, and the ability to converse with bots transforms the customer experience. Bots will use AI to manage unstructured data and complex tasks."
Facebook understands that bots can, and should, be more than just Chatbots. So bots are here, but the way we look at bots will have to change slightly. We are just now realizing the true potential of bots, and after a year we can see the conversational model might not be the only method we should be using.
But bots are absolutely necessary to revolutionize the customer experience and improve a contact center's bottom line. We just need to slightly tailor our expectations and understanding of interactions. Facebook has it right though: go where the consumers already are.