The Bots Are Coming
Customer communications is about to change forever, and you need to be prepared. Here's how.
The bot wars are beginning, and Facebook has just joined the battle, announcing Bots on Messenger at its developer conference in San Francisco today.
Facebook is not the first company to use bots for messaging. Among the enterprise communications players, for example, Slack has had a bot paradigm from day one, and Cisco and Microsoft recently launched their respective bot frameworks. But with a reported 900 million active users, Messenger has the potential to change business communications fundamentally.
What does it all mean for how businesses communicate with their customers? The promise of bot (sometimes called chatbot) technology is the complete automation of many business processes, from arranging a meeting to full customer service.
Think about planning a vacation today -- searching, comparing, booking flights and hotels, making payments, etc. -- lots of time and hassle! Imagine instead asking a travel agent bot inside Facebook Messenger to "check vacation options in Maui for June." From within Messenger, the bot has context of your previous vacation experiences, your airline preferences, restaurants where you've dined, and so on. The bot comes back with complete vacation package options, perhaps even including photos you posted from previous vacations in Maui to remind you of the highlights. If you like the options proposed you can pay right inside Messenger, and get your boarding passes there, too.
Behind the hype about bots are three real technology transitions. They are the:
- Explosion of consumer messaging apps
- Maturing of machine learning and artificial intelligence (AI)
- Emergence of conversational commerce
Each transition has implications for how businesses communicate with their customers -- and they are coming at us quickly.
Explosion of Consumer Messaging Apps
As I mentioned, Facebook Messenger now has a reported 900 million active users. As of January, WhatsApp had the same, while Chinese messaging app WeChat had 650 million, according to Statista. Messaging has eclipsed social media in terms of number of users -- and that's good news/bad news for you. The good news is that all of these users will soon communicate with your customer service reps via such platforms. The bad news is that if your business is not ready, they will talk to your competitors.
To get started, focus on your customer experience. Today's customer journey is often disjointed, leading to frustration for customers as they move among Web, mobile, and voice channels. A customer often has to start over in each channel, explaining who he or she is and what he or she wants to get done. The explosion of new messaging channels risks further fragmentation of the customer experience unless handled correctly.
If a customer sends me a message in Facebook Messenger today and in Telegram tomorrow, will I be able to continue the conversation meaningfully, or will I force him to start over? As you think about customer communications via consumer messaging apps, remember that consumers do increasingly expect businesses to maintain that single canonical, cross-channel conversation thread. So get started with a standalone Facebook Messenger pilot, but build a plan to deliver that consistent experience across the customer journey.
Bots and AI
So now that you've connected your people and customers using these new messaging channels, how about automating the conversation with a bot? Let's start with that word: "bot." While it sounds a bit scary, at its core a bot is a standard "if this then that" piece of logic. It's simply a piece of software that takes a specific action when it spots a particular word in a piece of text.
Here's an example of how we've used this internally. Everyone here contributes to our blog, and one of the fun jobs of our head of Digital Marketing is to gently chase our very busy engineers in Slack when a blog is due. Even our incredible engineers struggle to code with their right hands and blog with their left hands, so they built a simple bot that listens for the word "blog" in Slack and then plays this gif (inside Slack) of Woody Allen struggling to move his chair as he plays his cello in a marching band. It's their humorous way of keeping the demands of marketing at bay.
When actions are based on a constrained vocabulary, developers can program bots to behave as expected. Free-form natural language is harder to parse and analyze, but with machine learning bots can now be trained to get better at responding appropriately in free-form questions. The technology is still evolving though, as demonstrated recently by Microsoft's Tay chatbot, which Twitter users trained to sound racist and sexist in less than a day. Clearly this is not a risk a business should take with its brand, any more than it would do within a store or call center.
Before you unleash bots on your unsuspecting customers, do some trials of simple bots to automate employee workflows. If you are using Cisco Spark or Slack, for example, build a bot to automate a common task like inventory checks or opening leads in Salesforce.
This is the future. In a great blog about this concept, Chris Messina, developer experience lead at Uber, defines conversational commerce as "utilizing chat, messaging, or other natural language interfaces (i.e. voice) to interact with people, brands, or services and bots that heretofore have had no real place in the bidirectional, asynchronous messaging context." The idea is that business can be conducted end-to-end inside messaging apps. This is as much about a new user experience paradigm as a technology shift.
The evolution of user experience has come full circle. We started with the command line in DOS and Unix, evolved to the GUI, then to the Web and now to apps. One school of thought is that the new command line -- messaging -- will dominate, perhaps even eclipsing the browser and apps. Some look to WeChat as the example here, with billions of dollars in business transacted inside the app. A closer look reveals that much more than text is going on in WeChat, with pictures and HTML5 making it more of hybrid browser/messaging app. Facebook Messenger looks headed in the same direction, with support for structured messaging and call-to-action buttons.
The time to get ready for conversational commerce is now. Build a foundation by ensuring your brand can maintain a single conversation thread across the customer journey. But remember: Messaging is not social media and your organization should not delegate it to your social team!
Customers will expect to transact business and get things done inside messaging apps. Get started with bots soon -- first automate employee tasks and then simple, constrained customer conversations. When the AI technology is fully ready to go, you and your bots will be too.