The 4 steps for building the Future ChatBots

Nikila Kothari

Tweet “The future of chatbots” and watch what rolls in. To be fair chatbots are the future of every industry vertical today. They are everywhere. Every fortnight there is a new technology offering a new solution that are shoving chatbots into gadgets to make them more useful and powerful.

Most of the chatbots talk has been focused on Artificial Intelligence, and while it’s fine to build a chatbot, the chatbot is about so much more than a text conversation. Imagine you walking into a shopping mall. You look at the right and see Men’s collection. You look at the left and see the Women’s collection. Now close your eyes. Can you navigate your way to the kid’s section? Even though you are familiar with this shopping mall, without wayfinding and your sense of sight it’s unlikely that you could find your way.

Walking into a shopping mall with closed eyes is similar to most conversational interfaces today. Everything you need is available, but without the proper cues, information discovery is difficult.

So here’s where we are going to ask you to take a moment to think about what is a Chatbot?
You would probably answer with one of the two things – it’s a conversational robot, or it’s a computer program designed to simulate conversations with human users. But here’s the thing about chatbot: it’s not a thing. It’s not your smartphone, an app, or a killer robot. At its most basic, chatbot is a messaging application, sometimes referred to as a conversational interface, designed to simplify complex predefined task(s). Chatbots are very helpful when it comes to simple activities like ordering food, booking a taxi or check for movie tickets.

That said, chatbots are slowly becoming sophisticated as leading companies like Google, Facebook, Microsoft to Disney, Taco bell, CNN, just to name a few, are using them.

Let’s go down a typical path to building a solution with Chatbot.

1. Creating Identity for your Bot
A good bot should be meticulously thought out. Ask questions like: Why are you building this bot? Who is your target audience? What problems will the bot solve? How are you presenting the bot- as a person or as a robot? Etc. These are things you should precisely define before getting engrossed into the bot building process. Define the core use cases and stick to it.

Your bot should know it’s identity and it should understand its user expectations. When a user asks questions to a bot like “What is your name?” , “How can you help me?”, “What questions can I ask you?”, etc the bot must be able to respond instead of showing some static templated messages. Let us look at some examples of the existing chatbots:

CNN and PennyCat Chatbots
CNN is a news chatbot and PennyCat is a bot which helps you find deals for movie tickets, pizza, flowers and much more. In both these examples when asked for the bot’s name there is a templated message.

Users can interact well with a person whom they know and hence the chatbot needs to have a personality. You need to decide who your chatbot is and what matches its identity and actions. Like chatbot – “Hipmunk” has a persona: knowledgeable, Spontaneous and friendly. Likewise you need to define the persona of your chatbot and build an identity which reflects your brand instead of simply delivering information to users in an automated format.

2. Build your flow
Well structured conversational flow is the core to building effective and engaging conversational interfaces.

Building a Grocery Bot? Such bot can allow users to order groceries online and even automate their weekly meal preparation planning. The core of the conversation should be focussed on getting grocery to people. Many a times people go off rail, so you should always include the possibility for small talk. Your bot must trigger specific sentiments from your user then create hooks to leverage those sentiments and make the experience even better.

Conversation are limitless and probably your bot would not understand everything your users are going to want. Plan for this without breaking the flow. Even if your bot breaks, make the broken experience as painless as possible. To help you understand, let’s look at the below examples

ChatShopper and Allset Chatbots

ChatShopper is a shopping chatbot and AllSet allows users to book a table, order food and pre-pay meals to save the time. In both these chatbots the flow breaks when the bots are unable to answer some basic questions.

You should let your users know that your bot understands how painful it is and it wants to help you. Your bots should always drive the conversation forward. It should consider suggesting things to help users discover additional functionality. It should use actionable phrases or buttons to redirect them to a place that might solve their issue instead of leaving user with an empty white screen or an open ended question or statements that force them to decide what should happen next.

3. Content is king
The foundation of chatbot is content. Obviously you want your bot to be as human as possible. It must be something people enjoy coming back to. Use friendly, inclusive language when speaking to your users to make them feel they are chatting with a friend and they don’t feel like they are talking to a machine.

Does your bot work to fix its own errors? When your users are unhappy with the experience, will your bot recognize this and work to guide the conversation in a different direction or will continue on script, like emotionless robot? Let’s take up an example of the bot designed by CNN and ChatShopper.

CNN and ChatShopper Chatbots

They have a templated content which is displayed for any user response instead of replying with an emotionally appropriate response. Chatbots must save our time and reduce our efforts and pain. Otherwise, it’s not much better than any website or app.

While designing the content for the chatbots you have to create boundaries to conversation either by giving your users buttons to select or actionable commands to use. Don’t leave the conversation open to interpretation. Without boundaries your users can break your bot. Also be sure to repeat the info back so the users feel more comfortable knowing you got it right.

Like if I ask my bot to set a reminder for a meeting at 2pm, the bot should send me a confirmation message saying “I will send you a reminder for a meeting today at 2pm”. The Chatbot should be able to provide all the needful information within the messaging app making it a pleasant experience for the user.

4. Selection of Platforms
After coding your bot, you can start with the most exciting step i.e., selection of the desired channels! Facebook Messenger, Slack, Telegram, WeChat, etc platforms support bots. Select the platform based on your target audience and the user experience it offers.

What’s our take on this?
We think user interfaces should always simplify choices for the user, not add to their burden. We are moving towards an era where bots will be impacting customer experience. By presenting a few, relevant options, instead of an unlimited choice in the form of an open input field, the future chatbot should be built.

It’s time to create the Future of Chatbots
The 4 steps of building a chatbot – may seem like a lot of behind the scenes effort, but, they will all pay off when they come to create a chatbot that’s as friendly and likable as it is smart and efficient.

If you are looking for any help on building chatbots for any industry verticals, this is an area where we can help by designing a chatbot that is as charismatic as your brand.

SHARE

Leave a Comment