Digital technology constantly tempts us to do more with less. The promise of all these digital tools is that a single person can achieve what once took a whole team of marketers.
Each of us is immersed in a digital ecosystem consisting of hundreds of carefully curated and integrated apps and devices gradually built up over the years, until we don’t always notice just how many we come to rely on. Like an invisible Iron Man suit, this ecosystem superpowers our abilities in two important ways:
However, the allure of all this increased speed, efficiency and accuracy can lead businesses to devalue the human inside this metaphorical Iron Man suit. What else can technology replace? What other tasks can be delegated to the machine? Is the person inside the suit merely taking up space that could be filled with more technology?
Okay, maybe I’ve stretched the Iron Man analogy enough. But just because technology is extremely effective in some areas doesn’t mean we should seek to automate ourselves out of the picture.
In social media, one of the most obvious forms of automation is through the use of bots, which are basically automated social media accounts.
Well-conceived bots can be helpful and convenient in many small but significant ways. LA QuakeBot draws upon United States Geological Survey (USGS) data to notify followers of potential seismic emergencies around Los Angeles. Thread Reader is a bot designed to solve a specific problem for Twitter users, automatically unravelling long and hard-to-follow threads on request.
Meanwhile, other bots might be set up purely for fun, such as the Midsomer Murders Bot, which tweets randomly generated plots only marginally more ridiculous than those that appear on the long-running murder mystery series.
And then there are the bot accounts used by many media organisations to automatically share news articles as they are published. For example, the Twitter account of The Guardian is a fully automated feed of headlines and links to the latest stories on the news website.
People follow and interact with these automated accounts just as they might subscribe to a mailing list or sign up for SMS notifications.
However, these positive and benign uses are not what most people think of when the topic of social media bots pops up.
According to a 2018 study published by the Pew Research Centre, 66% of U.S. consumers are aware of the use of automated bots in social media. Of those, 80% believe bot accounts are mostly set up for malicious reasons.
Social media automation has developed a bad reputation. In recent years, the use of bots has become weaponised to promote certain political agendas, spread misinformation or simply to cause trouble – usually by attempting to deceive others into thinking these automated accounts are real people with real opinions. On the other end of the scale, some marketers have used bots to automatically generate (read: fake) higher engagement numbers on their posts.
This is automation as deception, which is why the social media platforms in general – and Instagram in particular – have been cracking down on such practices in recent years.
In short, good bots don’t hide that they’re bots.
In April 2012, Wired magazine spoke to Kristian Hammond, CTO of Narrative Science, about automated content writing. Hammond predicted that more than 90% of news articles would be computer generated within 15 years.
Eight years on, Hammond’s prediction seems a tad optimistic. Yes, computer-generated content is a thing – but it’s certainly not as widespread as the automation evangelists might have predicted.
Algorithms and AI are perfectly capable of matter-of-fact reporting, such as summarising a baseball game. The data are easily captured and strung together with full and fluid sentences, just as a voice assistant like Siri or Alexa does when you ask it for a weather report. Following rules such as word order, style and grammar is what computers do best, so automatically constructing sentences isn’t that big a deal when it is the external data that determines what each of those sentences is about.
However, a lot of writing doesn’t merely transcribe data into readable prose, requiring more human skills to determine what each sentence is about: imagination, perspective, empathy, critical thinking, cultural awareness, and so on.
Could this article be written by a machine? Yeah, nah.
I use this example not to warn marketers off AI-driven content – as I doubt most people reading this are genuinely considering it – but to illustrate how the desire to automate more and more of what we do can miss the point of what made the activity worth doing in the first place.
Automated push-button social media might create the illusion of effective activity, but the quality of the message is still far more important than the method used to send it.
Of course, it’s all about moderation and finding the right balance. Automation tools can help with routine and process-driven aspects of social media marketing – such as scheduling posts so that you don’t need to be at your desk whenever a message needs to go out.
But social media should never be allowed to run entirely on autopilot. Sure, it can save a lot of time while maintaining a steady flow of activity, but that activity would be increasingly one way. Too much automation risks turning social media into a broadcast medium, losing that community interaction that made social media so powerful in the first place.
Ultimately, the more a brand relies on automation, the less present the brand becomes. If your friends keep calling you to chat but only get the answering machine each time, eventually they’re going to stop calling. You need to pick up the phone more often than not to stop those relationships from decaying.
Similarly, when you’re hosting a community in social media, it’s only polite to actually turn up and join in the discussions once in a while.
After all, machines don’t build relationships: people do.