Choosing a Use Case for Your Bot

Episode #2 of the course Building your own Facebook Messenger bot by Carylyne Chan

 

Welcome back! Yesterday, we got a brief overview of the Messenger platform. Today, we’ll dive into one of the most important parts of building a bot: knowing what bot to build.

To be able to figure this out, think of what you do. Is the bot for your business or product? This is closely tied to what your objective of building a bot is. A few common objectives I’ve seen:

• Drive sales
• Support customers 24/7
• Increase brand recall and preference
• Innovate
• Have fun, as an experiment

These objectives inform the KPIs (key performance indicators) that you set for your bot to know if it’s indeed useful. (The only real exception is if you’re building a bot for its own sake, in which case you’re evaluating your bot by whether you could complete it well or not.)

There are several main categories of bots that you could build depending on what you intend for your users to do; for example, you can create a news bot if you’re a blogger, or you can build a bot to help customers shop if you own an online store. Productivity is another major use case; bot builders help users track their spending or habits or the weather, among other things.

If you are in a well-defined vertical, such as travel or finance, then you can mash together the functional use cases above. An example would be a personal finance bot that helps users log expenses while providing up-to-date promotions from merchants.

 

Features

After setting the general direction, brainstorm specific features that may be meaningful. You may want to build a news bot, but there are also features such as daily subscriptions, broadcasting, and chunking of articles that you have to consider. This is similar for other use cases too.

 

Resources available

Another important factor in choosing a use case is whether you have the resources to fulfill all the features that you want the bot to have. Scoping out the resources needed for a bot is much like how you would determine sprints for a mobile app or website.

For example, you may be interested in building an intricate, 200-outcome choose-your-own-adventure bot. Without the right team who can produce interlacing storylines, commentary, and choices (and money), it will be difficult to deliver on this specific use case.

 

Timeframe

Timelines are essential in planning for a bot. Bots are touted as simple replacements for apps these days, but that doesn’t mean they are any easier to build if you want to build them well. Creating a high-quality, robust, and scalable bot can take two months from start to finish.

Knowing your features, resources, and the timeframe you have will inform your decision on whether the use case you choose is feasible, and having a clear idea of the objectives will determine if it is eventually used by the people you intended it for.

 

Restrictions and possibilities

Finally, you have to consider the restrictions and creative possibilities that the platform may have. For example, in Messenger’s case, you are not allowed to send subscription messages unless you fall in categories such as news, productivity, or personal trackers. You are also only allowed a 24-hour window to re-engage with the user since their last interaction with the bot.

Tomorrow, we’ll cover the basics of planning and writing bot dialog flows, giving your bot a voice!

 

Share with friends