Part one in a series of articles about my experience with developing a chatbot using Microsoft Azure Bot Framework
Seems to be ages ago when as a kid I was scripting my first #IRC bot. We were able to create interactive text games, automate things, share files and even support a simple conversation. It was my very first immersion in a programming language and it definitely was a great learning experience.
I guess that life is made this way. Old things come and go and become the hot new thing again. Surprisingly enough, bots are no exception to this rule.
But don’t get ahead of yourself and try to reinstall your good old #IRC client! I’m talking here of the next bot generation: Smarter and easier to build!
There are a lot of bots out there for you to try
Whatever your allegiance is, if you use common channels such as Slack, Facebook Messenger, Telegram, GroupMe, SMS, or Skype ( just to name a few here 🙂 ), odds that you already had a conversation with a bot are pretty high.
If not, the best way to learn about bot is to try them yourself. You need to understand what they are and what to expect from them before building your own. There are a lot of bots out there for you to try. Basically, a bot for everything. An easy entry point is to browse bot listings such as the Microsoft Bot Directory or The Bot List.
For example, the And Chill chatbot is very useful when it comes to answering a simple question: “What are we watching tonight?”. The bot will suggest movies you might love to watch based on movies that you already viewed. No more endless conversations with your mates to choose a movie; this bot solves it all.
If these types of bots are not your forte, you might love something more practical such as the Subway Order Bot that lets you preorder your beloved sandwich for pickup at your nearest Subway location. Now imagine how easy it is to take the order for your party since you don’t have to remember everyone’s sandwich and just let the bot handle it for you. Similar bots to visit are FreshDirect, The Cheesecake Factory, and the Domino’s Pizza bot.
Bots are also good for all audiences, large to more specific niches. For example, Netflix built a Slack bot called hubcommander. The bot creates a simple way to perform privileged GitHub organization management tasks without granting administrative or owner privileges, which saves organizations both time and money. Want to know more about this bot? Visit the Netflix GitHub Repository.
Business opportunities are booming
“We were looking at the numbers, and there’s around 60 million businesses on Facebook, which means there’s a lot of opportunities for businesses to create tools to better reach customers,” says R/GA San Francisco executive creative director Paulo Melchiori.
There was a time when social media was considered by some as a passing trend; however, at this point in the game, not having an active social media presence is like pulling out a flip phone. No matter what you sell and who you sell it to, using social media as a marketing tool is inevitable.
With around 60 million businesses on Facebook and with more than 1.94 billion monthly active users who spend around 50 minutes each day on it, business opportunities are booming for the bot industry.
What problems do chatbots actually solve?
One of the items rising up from the customer experience agenda is the objective of creating a frictionless experience. Bots, by their nature, bring a more natural flow and eliminate friction between you and your users.
For example, asking clients to download and install your app could be the line between having a new user or not. Bots are solving this issue and make the onboarding easier since they are relying on commonly installed channels, such as Facebook. To put things in context, Facebook is the most installed app right after the weather app.
Another common friction point is account registration. Bots make things easier and let your users link their existing social media accounts with you bot. You will then have access to a unique ID that you could associate with an account on your end. No need for your users to create a new account, or learn a new password. Everything is done in the background and seamless for your users.
You need to remember your client from conversation to conversation; don’t make them tell you things you already know. Bots are actually very good at this since they keep the context of all your conversations. This way you can refer to any information you’ve stored in the context and reuse it. For example, you can ask the user’s location and reuse it every time they do a search and filter the results based on this information.
“You want to listen what your customers say and reach out with a more personalized approach.”
This where the fun happens. What could be more personalized than a direct conversation with your client? Bots, more specifically chat bots, are made for handling a conversation. They understand the intent of what your client want and respond accordingly. They have a personality representative of your target audience and help your clients achieve their task at the moment they want. Pure frictionless!
KLM on Messenger is a great example of a personalized conversation.
Sidlee (one of the largest ad agencies in North America) recently used Facebook Messenger to host a ludic bot for a business named Skyn, a condom and lubricant manufacturer. Using the bot as a new channel, they reached their audience by offering them a new personalized experience. The Skyn bot is a voice analyzer that lets users know what their mystery factor is. I’m no Barry White, but my mystery factor sure did score high. 🙂 Give it a try yourself!
What is next?
In part two, we will be more hands-on and talk about the developing process of a chatbot with Microsoft Bot Framework.