Facebook and Twitter are two of the biggest and most high-profile social networks in the world, but they’ve taken very different approaches when it comes to handling bots on their platforms.
In this post, we’ll look at the ways bots are changing both platforms, as well as the challenges both have faced and what they’re doing to address them.
A quick bot refresher
Remember that when we talk about bots we’re (generally) talking about relatively simple web apps that use some kind of chat or text interface for their UI layer. In the case of Facebook Messenger and Twitter, bot makers use a developer API to connect their bot to the service.
Beyond that commonality, what these bots do and what role they play on their platforms are as different as Facebook and Twitter themselves.
How bots work on Twitter
Twitter bots have been around since at least 2010, when the first malicious bot attack occurred. (As if to foreshadow what would come later, this first attack was part of a particularly nasty political campaign.) Twitter bots are accounts just like any human account–they can tweet, like, follow, unfollow, and DM other accounts.
Because from the outside they look just like human accounts, it’s impossible to say exactly how many Twitter bots there are. A survey from March 2017 suggests there could be more than 48 million bot accounts (that’s out of around 330 million total monthly active users).
The Good, the bad, and the pointless
Many of these accounts are benign, fun, or useful. The Netflix bot tells you about new movies and TV shows available for streaming on Netflix in the U.S. Medieval Death Bot shares random death records from medieval coroners’ rolls and automatically generates a suitably medieval death for you if you tweet at it. (Spoiler: Most likely you’ll be killed by clerks.) Dear Assistant is a general Q&A bot hooked up to the Wolfram Alpha answer engine.
These aren’t the kinds of bots Twitter is concerned about, though. The vast majority of Twitter bots seem to be devoted to inflating follower counts and spamming links to articles. In fact, a Pew survey estimates that as many as two-thirds of all tweeted links were shared by bots.
The buying and selling of Twitter followers is a multimillion-dollar business, one that Twitter fears could undermine it as a platform. As with the streaming service Twitch, advertisers may be less inclined to buy ads on a service where a large percentage of its “impressions” might not even be human. Beyond that, businesses that make decisions based on social media presence may be less likely to use Twitter activity as a valuable metric.
How bots work on Facebook Messenger
Facebook Messenger launched an API for bots in 2016. These chatbots tend to be more sophisticated than your average bot–for example, Facebook’s API allows developers to add app-like functionality to their bots.
The robust functionality available to bot developers, as well as the ubiquity of Facebook itself, makes Messenger an attractive platform for brands looking to throw their hat into the chatbot arena. CNN’s daily news bot serves up the latest news stories but also allows users to customize it to make sure they get only the updates they want to see. Niki.ai is a bot that lets users order cabs, book hotels, order takeout, and even pay their utility bills right from Messenger. Duolingo lets users practice their foreign-language skills with a wide assortment of customizable bots across a number of popular languages.
Mo’ data, mo’ problems
Facebook’s bot problems have less to do with the way the bots are used on Messenger than with Facebook’s privacy woes in general. Following the Cambridge Analytica data breach that exposed the data of more than 87 million users, Facebook promised to thoroughly review the way it handles users’ personal data on its platform.
While the data breach didn’t directly involve the Messenger platform or bots, Facebook announced that it was putting a moratorium on approving new bots for Messenger until it can conduct a full review of the platform. What this review will mean for Messenger bots going forward remains to be seen.
Want to learn more?
See how chatbots are changing social networks like Discord and Twitch. If you’re interested in learning how to build a chatbot, whether for any platform, check out our article about making chatbots part of your revenue model or engage a chatbot development specialist today.
