You may have heard of bad bots which spam people or make fake accounts. However, not all bots are bad. There are good chatbots and legal chatbots as well. One of the most popular bots is the Messenger chatbot which uses the official Facebook API to communicate with fans, clients, and anyone on Facebook or Messenger.
What are Messenger chatbots?
If a business has a chatbot installed on their Facebook page, when a person starts a conversation with the page, the chatbot will respond automatically to the person.
Chatbots can be programmed to reply to trigger words with a certain message or chat sequence.
For instance, if a person asks a pizzeria's fan page "What are your opening hours?" the bot can be programmed to recognize words such as "opening hours" or "when are you open" and then reply with a relevant message.
Messenger chatbots can include images, videos, audio recordings, sliders, and buttons/links to webpages.
How can businesses use Messenger chatbots?
A chatbot can be used to answer common questions people often ask a business on Facebook. People hate waiting for a reply and if no human support person is only to answer them, they can get very annoyed. A chatbot is always working, always online, always there to answer people.
Messenger chatbots can also be used to grow a list of subscribers who can be messaged in one go (with none promotional content only!) at a later time.
You can control how and where your bot's messages will appear, if they'll make a tone as they are received, if they'll appear on the person's home screen/lock screen of their phone or just in their list of conversations in Messenger.
In theory, Messenger chatbots are a really awesome thing with many opportunities and possible uses. However, in practice, people do not always respond favorably to them.
What are the benefits of using Messenger chatbots?
- Your bot can reply to people's questions at any hour of the day or night.
- You can have a default reply/menu with buttons, links, and images showcasing your products.
- You can grow a list of subscribers whom you can contact again later.
- Depending on yoru niche and country, you could grow a list of subscribers for cheap when running Facebook ads. There are many options how to get peopel subscribed to your bot. For instance, you could promote a post on your fanpage and whenever someone comments on that post, the bot kicks in and asks them a question or gives them a link to a freebie/a coupon. When they reply to the bot, they become subscribed. In some cases, I was able to obtain a subscriber for only $0.02 dollars which is much cheaper than pay-per-click ads. (In other cases, a subscriber cost $1.20 to $3.00.)
- Best of all, the moment you have a bot installed on your page, your page can be submitted to the Messenger bot directory and be discovered by users searching the directory for bots!
What are the dangers of using Messenger chatbots?
The Facebook API is sometimes very unreliable and suffers downtime. If the API has issues, messages from your bot don't get sent or arrive hours late, buttons in the bot don't work, or the bot sometimes goes crazy and sends the same message 10 to 100 times (which drives the user crazy).
In some cases, using a chatbot can seriously damage a brand and cause the business to receive thousands of bad reviews on their Facebook page. Here's why.
 People fear bots or feel insulted a bot is talking to them.
Although some companies claim to have "AI bots", currently all chatbots react to trigger words. If a bot isn't set up properly, or a person misspells a word, the bot doesn't know how to respond and could either reply with a very annoying irrelevant message or just with the default "Sorry, I don't understand" message.
Some people use FB chatbots to collect email addresses or personal information. The problem is, most people are on their phone and often misspell their email, name or other information. IF the person misspells his email and then realizes it, and tries to say "oops I misspelled that, this is my email" but the bot already is asking him for his phone, the bot will get confused and not know what to do (unless the developer spent a lot of time making complex logic flows and triggers).
Additionally, when the bot tells people it's a bot (which it is obliged to), many people don't understand what a bot is and get offended and annoyed that a bot is talking to them and block the page or report it. Others are sometimes terrified of thew ord "bot", scared a nasty bot will do something evil to them (hack their account or steal their personal data). While others don't understand it's a bot and keep trying to chat with it and ask it questions which it doesn't know how to reply to.
In my experiments with building Messenger subscriber lists across different niches, about 10-20% of the people immediately unsubscribe when they hear it's a bot ore reply with angry messages.
 People don't know how to unsubscribe. A bot could irreparably ruin your brand.
Because the bot looks like a human chatting with the people, many people treat it like a human and think it will understand what they mean.
Even though the bot tells them to type "stop" in order to unsubscribe, most of my subscribers came up with infinite other ways to say they want to unsubscribe.
"Leave me alone", "Please quit messaging me", "Enough already!", "How do I get off this thing?", "You're annoying. I'm reporting you!", "Leave me alone or I'll call the police"
And those who did use the stop word didn't just say "stop" but many different versions of "stooooop".
Often even when the bot would reply with a default message of "text "stop" to unsubscribe", the people would just reply with a "yes".
Then, when they got angry enough, they'd report the page to Facebook or they'd go writing angry reviews and posts on the fan page (which can really damage one's reputation and ratings).
 It's hard to ask people questions in a bot.
Some businesses use chatbots to pass people through a questionnaire (for quizzes or for real estate inquiries).
Unfortunately, from my experience, when a bot asks someone "Do you want a house or an apartment?" many people simply reply with "yes" instead of pressing on the appropriate option or button.
In most cases, a web form is an easier, faster option with fewer mistakes and less annoyance to people.
Nonetheless, Messenger chatbots are very promising and when people grow more accustomed to bots and the bots become smarter, they will probably be fewer problems using bots.
My advice for businesses who want to use Messenger chatbots is to use them with caution. Carefully monitor the live conversations and step in as soon as the bot makes a mistake or annoys a person, before the person gets too angry and leaves a bad review.
If you're using a 3rd party platform like ManyChat, set up an alert (on any free website uptime monitor service like uptimerobot.com or siteuptime.com) to see when their website goes down so you can immediately rush to Messenger to manually answer people's messages. Several times in the past during my advertising campaigns, their website went down for 30 minutes to an hour or more. If their website is down, none of your bots will work.
What is your experience with Messenger chatbots?