Using Chatbots to Improve User Engagement (Part 1 of 2)

This is part one of a two-part series that takes an in-depth look at how Progress Labs is innovating with Chatbots. You may read part two here.

Chatbots and AI are rapidly developing and powering a new wave of customer engagement. It’s not too late to pull ahead of the crowd, and Progress can help.

Do you ever ask Google Now to send a text message to your spouse while you’re driving? Or do you ever go to a new city and ask Siri to help you find the perfect coffee shop? These are just everyday examples of how bots are improving user experience. This blog post will introduce companies, big or small, to bot technology—and offer suggestions on how they might be able to utilize it to engage their users in the best possible way, rewarding them with an unprecedented experience.

Progress Labs is currently exploring bot technology and the various ways that it could potentially deliver business benefit for our customers and partners. While we don’t directly offer bots as part of our software, the technology exists today to build bots with the help of Progress OpenEdge.

The best way to prove an idea or a concept is to do it yourself. We created several chat bots for businesses and kept it simple, so that anyone could build a bot even with limited knowledge and experience.

What Exactly Are Chatbots?

Chatbots are software programs capable of establishing conversations with human users, preferably in their natural language (but not necessarily limited to text conversation). They can also facilitate voice, virtual reality, and other means of communication. Two interesting applications of Chatbots include:

Chatbot Lawyer

This bot helps users file legal claims for parking tickets and other simple petitions. To achieve this, it asks multiple questions and comes up with an appeal letter. Once the appeal letter is generated, it can be directly mailed to the court.

bot-lawyer

The bot is useful in this context because laws are publicly available and the trivial tasks which were once accomplished by a lawyer decades ago can now be automated.

It can also help users with filing insurance claims in case of delayed or canceled flights.

You can access this particular bot and read more about it in this article.

The TacoBot

TacoBot is a Slack bot from Taco Bell. It takes orders via chat and helps users with related queries like item customization, order quantities, and total cost:

tacobot_preview

TacoBot is still in private beta, but sounds like it will be made available soon.

Why Should You Care?

The above examples make it obvious that businesses are transforming and they are looking to use chatbots as the primary way to engage with their users. But there is good news! It’s not too late – the train has yet to leave the station. Most of these bots are in private beta mode and everyone is making earnest efforts to put them into public use. It is still in the early stages, so you can be ahead if you choose to put your focus on bots. Here are a few statistics to support the claim for the future of bots:

Be it ordering food, booking flight tickets, filing legal claims, launching a personalized survey and so on, every business can find a need for a Chatbot to improve the user experience.

How Do I Create a Bot for My Business?

Building a bot for yourself is not the tough part, but deciding what you want the bot to do can sometimes be confusing.

Here are the top seven questions you should ask yourself before writing a bot for your business:

  • What primary customer use case is your bot going to address?
  • Who are the primary users of your bot?
  • What channels you want to support?
  • What are the actions your bot can perform?
  • Is there a need of manual intervention at certain point(s)?
  • How deep and nested do you want conversations to go?
  • How would you like to handle unexpected conversations?

These questions are important to define a clear specification for your bot. The Chatbot business is fairly simple but can get messy if the conversations go uncontrolled and unnecessarily deep.

At the same time, it is important to define the channels you want to communicate. In the beginning, supporting all available channels might appeal to you. But there are reasons to avoid multiple channels:

  • Only a few channels, like Facebook Messenger and Skype, offer you advanced controls like cards and calendars
  • It’s much easier to maintain and debug fewer channels

Once you have some clear understanding of the requirements of your bot, you may elect to write one. Writing a bot is not just coding, there are few more phases to it. In fact, coding is probably just about 25% of what you do. It also involves training the bot, defining response templates, maintaining the conversational infrastructure, integrating it with messaging channels and the backend system of records.

Thankfully, the bot framework provided by Microsoft offers most of these out of the box and requires you to only write the brain of the bot.

Developing Chatbots Using the Microsoft Bot Framework

As a part of this Progress Labs activity, we built two bots from scratch:

Warehouse Bot

The first is a warehouse bot – it is capable of answering questions related to warehouses and stock levels. For example, you can ask for a sales report, the status of stock in a warehouse, and configure it to notify you about depleting stock items.

bot-warehouse

Our Chatbot demonstrates the capability to serve rich media content like Kendo UI charts and graphs, images, and HTML. The Chatbot can be integrated with various channels like Skype, Outlook, SMS, and many others.

Notification Bot

Another proof-of-concept we built is a proactive notification bot. It notifies relevant users about a new order being placed and acts upon it based on the response from the user. What you see here is an SMS conversation initiated by the bot:

bot-sms

Again, the channel of communication can be SMS, Skype, Slack, Outlook, and so on. This POC demonstrates that it is not always necessary that the conversation be started by a user. Sometimes when the bot feels necessary, it can start its own conversation.

What’s Next?

Chatbots look fantastic, are fun to interact with, and give you the power to establish meaningful natural language communication with your apps. But the beauty of bots lies in the simplicity of the system. In the next post in this series, we will dive into the code and see how easy it is to understand bots via a handful of components. Continue to part two here.

Header image courtesy of Eelke

Comments