Earlier this year, I introduced Facebook Chatbots in one of my blogs. It was part of a promotional overview we executed for a client. To reiterate:
Enter the clever AI of Facebook Messenger chatbots. Chatbots can be used to automate conversations. The design can be as simple as answering a few questions and including a few links. Alternatively, it can be elaborate workflows that span e-commerce options. Most importantly, it’s a gateway for your audience. It’s a space that’s tailored just for them, and one that can supplement their journey, be it research, purchase intent, or ready to make a decision.https://www.acmarketingcaribbean.com/what-i-learned-from-using-facebook-chatbots/
Before said chatbot, there was Sam, ACM’s Content Manager. Her role within the communities we manage went beyond simply replying to comments and direct messages. Her online presence morphed into a trusted source, a confidante, a reliable communicator. That consistent and personalized two-way communication gave our communities confidence in engaging with us.
When we first introduced a chatbot into our day-to-day workflow, our major concern was maintaining said personalization. Sam and her accompanying personality were staples of those smooth online exchanges. How could that be translated into an artificial intelligence (AI) environment?
There are many common misconceptions of community management on social media. Sam can probably write a book (and then some) on the subject. Introducing an additional layer of interactivity for our audience in no way meant compromising on the quality of said interactions. As I wrote:
Our digital immersion has not taken away from the psychological need to feel acknowledged, befriended, and cared for.https://www.acmarketingcaribbean.com/what-i-learned-from-using-facebook-chatbots/
This was (and is) our general consensus in creating and managing how people connect with the chatbot. In the months that followed, we’ve learned the following:
Reproducing Natural Speech
There is a marked difference in “Hello” and “Hello, Nicole”. With the addition of multiple layers within the chatbot, that individualized feeling needed to be maintained. For example, a natural conversational flow can be achieved through simple tricks such as strategically placed pauses (as shown above). Automation is a powerful time-saver. However, an organic (as opposed to robotic) approach should not be taken for granted.
Tweaking Based on Audience Needs
Through our maintenance content, our audience provided us with insight into their frequently asked questions. Through this, we were able to create specific choices within the chatbot. As I’ve written before, it’s all about your audience needs. So, we deliver useful information beyond promotions and campaigns. This decreases the need for human intervention while also satisfying our audience’s behaviour.
You may be aware of the many discussions around Facebook privacy and data. With that in mind, it was important to give our audience the choice of being part of the chatbot’s broadcast list. Every time someone interacts with the bot, they are given an opt-in/out option. This subtlely ties into the first two points. Once the bot is effectively designed and the information remains relevant, the audience will willingly give their permission.
Are you familiar with Facebook chatbots, have spent time testing them or have interacted with brands that use them? What has your experience and learnings been? Share them with me in the comments below.
Subscribe to ACM Weekly for insight into content, data, and advertising.
Register here to be notified about ACM’s monthly LinkedIn Local events.