A Marketer’s Guide to Facebook Messenger Bots

When Facebook announced that bots were coming to Messenger, it wasn’t a surprise. It was mid-April of 2016 and popular apps like Kik, Telegram, and a few others were already successfully experimenting with the technology. Still, Zuckerberg’s announcement was a welcomed expansion among the developer community. They weren’t the first to market, but they were easily the largest player in the industry. Facebook, the world’s most popular mobile app, has brought a completely new channel of communication to its users.

With any new technology, there are always lots of questions mixed among the opportunities. It has been a few months since the public release, and digital marketers and developers have been scrambling to take advantage of the new tech, some more successful than others. Since then, we have seen bugs get sorted and the user experience improved dramatically.

The technology is currently primed for massive growth. The adoption curve just now hitting a critical mass, businesses around the world are beginning to engage their users through Facebook Messenger bots. In a world full of technology, often feeling stale and static, bots are friendly and natural, providing information in a way that no other digital medium can match.

How does a Facebook Messenger bot work?

It starts with Facebook Messenger, the companion app to Facebook that provides a platform for Facebook users to communicate with each other. In Messenger, users can send GIFs, videos, links, and even share events and updates from their Facebook News Feed directly into a conversation. Today, Facebook’s chat platform is on over 75% of mobile devices in the US, and with each new feature, Messenger continues to become more powerful.

The technical side is quite simple. Bots work by creating structured paths and journeys for users based on rules to deliver a catered experience. For example, a retailer looking to improve top-of-mind awareness and online sales could deliver weekly newsletters or create an automated shopping experience. Building an eCommerce platform on a bot allows people to shop at their convenience on a platform they have been using for years. Like most marketing efforts, the experience focused on getting users to take specific actions. There are no distractions from competing ads or unrelated content pulling people away from the shopping experience. One successful example of this is the New York apparel retailer, Spring.

Using Spring’s Messenger bot, users can browse selections, filter to styles and sizes, find similar items, and ultimately make a purchase delivered right to their door. All of this can be done in a matter of a few minutes using a Messenger bot. We tried it out. You can see our example below.

Messenger bots are not absent of challenges

There are some challenges with the current technology. Through the guidelines developers establish, a bot will always respond with a predetermined answer. Additionally, the novelty of a bot can wear off. Developing a useful Messenger bot will take time and a strong strategy. But depending on the structure, a bot can act both as a piece of content marketing or a typical outbound advertising medium. 

Bots are also isolated from each other. There is no general bot you can chat with that will understand unrelated questions and be able to direct you to the appropriate solution. For example, you couldn’t ask the Yahoo! Finance bot about specials at a local Domino’s and expect a useful answer. Most bots function by identifying certain keywords to determine a user’s intention, limiting the types of inputs to specific commands. Even phrases like “Hmmm…” or “What about” convolute the message enough to confuse the bot, even if the command is obvious to any human. In our example below, Wall Street Journal says I can look up any stock by entering its ticker or by saying “Lookup ‘company name,’” but by simply adding “Hmmm…” prior to my answer, the bot doesn’t understand my command.

In this example, notice how the Wall Street Journal bot struggles when I use words outside of its direct command language.

Despite the challenges, Facebook is currently dedicating millions of dollars to improve the capabilities of a natural language AI. For now, things are still pretty limited.

The capabilities are endless

Even with all of these barriers, the solutions seem endless.

Take emergency communication systems for example. Rather than public Amber Alerts or SMS, limiting messages to text or audio, imagine receiving images, maps, step-by-step directions, live alerts, and videos during emergency situations. Or, what if public transportation networks could alert commuters of slowdowns or detours while providing alternative lines or maps in real-time? Retailers could offer tailored coupons to their most dedicated fans. Content marketers could reach new heights through location and interest-based solutions. Record labels could send new albums and recommended artists specifically to attendees of concerts and events.

The potential reach of Messenger bots in any business or vertical is staggering.

Bots are a prime opportunity for innovative marketers

For decades, digital marketing and technology have been plagued by its lack of human character and emotion. Computers lack empathy. They don’t understand feelings, emotions, or needs. While the bot is only an early step to creating a seamless computer-to-human relationship, it should not be overlooked. There are challenges, and we are solving them quickly. In time, more of our interactions with computers will appear to be a seamless extension of our personal lives. For now, Facebook’s Messenger bot is a tech-forward way to give your customers powerful and relevant content through the most popular social platform in the world.

Check out two more demos showcasing newsletter features and informative content.

Up Next

Will the Presidential Election Affect Consumer Spending?

It’s a question that business owners and marketers have been asking. Will the outcome of our US presidential election affect…

John Drake
John Drake
August 25, 20161 min read