Facebook Messenger bots (chatbots) are a controversial feature that was added to the standalone Messenger service earlier this year. Recently, Facebook started allowing its Messenger chatbots to take payments – making the interaction between chatbot not only a customer service feature but an e-commerce feature. Jes Stiles of Search Engine Journal writes that “Facebook Messenger is the most attractive platform on which to run bots for marketers…due to 1 billion monthly active users, payment, and advertiser integration, as well as a healthy developer ecosystem.”
Over the course of this year, chatbots have been deployed on the platform in some very interesting ways:
- The New York Times developed a chatbot that updated users about key aspects of the election, automatically answering (some) specific questions about the election
- Estee Lauder is using Messenger bots for a mascara sales campaign
- Food Network developed a Messenger bot to help to find recipes with specific ingredients
Note that a chatbot isn’t the same as an autoresponder. Wikipedia discusses that ” Some chatterbots use sophisticated natural language processing (NLP) systems, but many simpler systems scan for keywords within the input, then pull a reply with the most matching keywords, or the most similar wording pattern, from a database.” Chatbots are more elaborately planned and more functional than an autoresponder. Chatbots trigger different content and logic based upon user input, and there are contextual tools that allow Facebook Messenger bots to parse language deviations.
I think one of the cool aspects of this technology is that it’s an opportunity for a smaller business to scale a bit of their customer service; to perhaps appear bigger than they are. And Facebook is building up an analytics tool to help create and track metrics around Facebook Messenger bots. What I want to do is to explore the ways that you (YOU!) can implement a simple Messenger chatbot into your social toolbox (somewhere on the less sophisticated spectrum).
The third-party chatbot
In this category of Messenger Chatbot, I include some customizable chatbot solutions that don’t require any coding on your end. These tools demonstrate that it is relatively straightforward and inexpensive to deploy Messenger Bots, but also that there is some elaborate thought that must go into making these tools useful for visitors.
Flow XO (flowxo.com)
Flow XO is a user-friendly chatbot development tool that allows you to develop keyword-based bots in Messenger (as well as other platforms such as SMS and Slack). The advantage of FlowXO is that it is very user-friendly and has an active community of users that ask and answer questions and share best practices. The bad news is that despite not having to code, the community posts demonstrate that developing an effective bot requires pretty sophisticated planning.
Botsify has a similar user experience to Flow XO – no code needed, very straightforward user interface. Its selling point appears to be integration with WordPress and Medium (building a bot that points users to other relevant site material could be pretty useful for content creators). The Botsify interface is more straightforward than Flow XO – everything is oriented towards the bot “conversation” that you want visitors to have. And the step-by-step instructions on Botsify are spot-on and easy to follow. While it appears that Flow XO has more options and integrations, Botsify may be the easier third-party bot builder to navigate.
Pricing for Botsify is free for a limited account and paid accounts start from $10 / month.
Chatfuel is a dashboard-driven bot creator that is low-cost (free), allows testing during development of the bot, and (!!!) allows payments. While the Chatfuel dashboard isn’t as user-friendly an interface as Botsify – Chatfuel’s quick incorporation of the e-commerce opportunity may be an important feature for some businesses.
Chatfuel is free up to 100,000 conversations a month…an unsustainable monetization model, but a great price (for now) to get started with Messenger Chatbots.
Converse is a multi-platform chatbot creator (of course one of these platforms is Facebook Messenger). What is pretty unique about Converse is that you set-up your bots in a flow chart rather than with the guided menus featured in the other solutions. For people who prefer a visual representation of cause and effect, this may be an ideal solution. It has robust help documentation but perhaps is a little more difficult to set-up than the other solutions.
Like the other apps, Converse has a free option that escalates in the next tier to $150 per month.
Perhaps similar to Converse, api.ai has a very programmatic feel. Organized into “intents” and “entities,” api.ai offer some opportunities to perform very elaborate activities using Facebook Messenger Bots (the video below demonstrates how to set up online pizza orders with a multitude of variables). The interface is perhaps the least immediately intuitive of the apps described, but the support documentation is fantastic and enables you to perform some pretty advanced things with your bot. Also of note, Google now owns api.ai.
The cost for api.ai is free for limited bandwidth and requires that you “contact sales” for its paid product.
The “I know how to code a bit” chatbot
If a comprehensive solution doesn’t strike your fancy and you’re comfortable with coding, there are quite a few demonstrations of how to make Messenger bot by your lonesome:
- The Facebook Messenger team has a pretty elaborate set of examples of how to code your Messenger bot (a more detailed demonstration of the specifics can be found here)
- Hartley Brody published a detailed demonstration of how to set up Facebook Messenger bots here.
- Google or search on YouTube or Github to find lots of examples of how you can code your Facebook Messenger bots.
Bottom line: if you have the knowledge and diligence to code your own Messenger bots (self-disclosure: I have neither) you will be able to utilize the complement of tools that Facebook Developers allow for Messenger bots.
You can make a Facebook Messenger Bot. That’s amazing, right? For the very low price of free, you could sell things online using parameters and payments initiated by a bot on Facebook Messenger. You may believe that chatbots are a paradigm shift in human interaction.
If this exercise demonstrates anything for you, hopefully it’s that chatbots are a (sometimes) elaborate tool for you to add to your social toolbox. They’re only as smart as the logic that you put into them, and as Twitter Bots have demonstrated, the logic that you program into them has a reasonable probability for (epic?) failure.
That said, there is potential for Messenger bots to add value to your social, customer service, and sales strategies. Now that you understand how easy it is to implement Messenger bots with third-party tools, perhaps you’ll use this social strategy in the future.
(Photo Credit: Pixabay)