As the number of AI-enabled products grows, our lives are increasingly influenced by robots. A great proportion of these“influencers” are Social Bots. There has been a lot of debate recently on whether social bots are changing the way we think and whether social media accounts run by bots have any influence on our beliefs and opinions.
What has happened to people?
If you’re from the UK or the US, you have almost definitely spent over two years being inundated with news on two persistent, ever-present topics: Brexit, and Donald Trump’s reign as president.
These two events, a referendum and an election, have been at the forefront of political news for years now due to the controversial nature of both, and the great rifts they created within the respective populations.
Rarely throughout history has there been such a sense of division throughout the British and American people.
Rarely has there been such vociferous debate from all corners and rarely has there been so much noise, so much frustration pouring out from the masses, directed at the other 46%, the other 48%, the other 52%.
And where have we seen all this noise, all this discontent, all these arguments? On social media. Comment sections of political news posts, articles and stats have become swamped with users sparking fierce debate.
They have become a haven for the opinionated ‘real person’ to vent and rant and push their views. For good reason, because it’s a forum that is accessible to anyone with a computer and allows anyone to get their views in front of a lot of pairs of eyes.
Sometimes you find yourself barely believing the keyboard fury in these dark spaces of the internet, and start asking yourself, ‘what is wrong with people?’. Well, what if these people, aren’t actually people?
What if social opinion, so impressionable and so influenced by material we read on screens, was actually swayed by the polar opposite to the ‘real person’?
What if in fact, the way we vote, and the way we think, is being influenced by just a few people with some serious outreach. Well, that’s the path social bots are leading us down.
In this post we’re going to give you the full lowdown one what social bots are, how they work, why we should actually be a little worried about them, and how you should use them as a marketer.
What are Social Bots?
A social bot, put simply, is an automated social media account, that isn’t representative of a human being in the real world.
Originally, these bots were designed to simply take on repetitive or menial tasks, but as machine learning has advanced, they have become far more useful for those with an ulterior motive.
Do you ever read a comment that looks like it’s been written in bad English, is serving simply to divide opinion, ‘trigger’ or ‘troll’ other users, and uses the same couple of words or phrases multiple times? The odds are, you’re looking at a bot.
As they advance though, they can be difficult to spot. In fact, it’s thought that 30% of social media users can’t actually spot the difference between a bot and a real user.
This means there is a level of trust put in these bots that is completely unfounded. What people often think are real views and educated opinions, are often nothing more than a piece of code responding to an algorithm.
They are generally created by one social media user, or botmaster, who will make a huge number of automated bots to go out and spread whatever message they programme them to.
Studies have shown that it can cost as little as $100 to assemble one of these botnets, and take little under half an hour.
Botnets, all with the general purpose of silencing real debate, real views, real opinions, with their own clutter, can be targeted to talk about a certain topic, hashtag or issue. More often than not, they succeed in silencing these debates.
Botnets are made with tens, if not hundreds of thousands of individual bots, and easily dilute the gauges used to determine opinion, popularity and the legitimacy of news stories.
Single bots have been shown to have far more retweets and shares for example, that any human user, due to the fact that they can tweet far more regularly than human users.
A bot doesn’t need sleep, and can post or tweet relentlessly throughout the day, or alter their patterns of behaviour to mimic a human user, posting at certain times of day.
The upshot is that misinformation is spread rapidly, human users are lost in a haze of repeated mantras and false news, conspiracy theories and twisted facts.
Does it work? Well let us relay you some of the facts about the 2016 US presidential election, a definitive moment in the history of politics that ushered in (along with the Brexit vote) a new, social media-driven era of politics.
The day before that infamous result, that elected one of the most divisive presidents ever, a study was released that said that there were 400,000 bots on Twitter alone, driving a staggering 20% of political chatter.
One in every five tweets, retweets or shares, was the work of a bot. A lot of the time, they were spreading misinformation and divisive material.
Often, the bots simply release the misinformation or the sensationalist, extremist material, then let human users go to work on sharing it.
Through this method, false information can reach millions of people as users see their friends and family sharing these posts. The result? Influence over millions of votes.
Should We Worry About Social Bots?
We’re sure you’ve heard the stories about the Russian government being involved with Trump’s election, and unfortunately, thanks to the power of Facebook, they were.
One year after the election, Facebook revealed that Russia’s ad campaign had reached 126 million Facebook users from early 2015, to mid-2017.
Now, if you consider that the US population is estimated at around 325 million, if these figures are accurate, then that would mean that almost one third of the entire population of the US had been reached with Russian propaganda.
Facebook’s ability as a platform to target users with particular ads by analysing thousands of characteristics is, unfortunately, ripe for abuse.
This highly targeted form of advertising allows organisations such as the Internet Research Agency to carry out campaigns that reach just the right people with divisive content.
Almost 62 million people voted for Donald Trump, that’s just under 20% of the entire population of the US, twice as many were subject to ads from the Russian government.
Does this mean that people were actually affected by all this marketing though? The simple answer is, yes.
How can you tell it worked? In a study conducted by the Oxford Institute, that compared the social bot campaigns done by both the Clinton and Trump campaigns during the 2016 presidential race, the results were somewhat telling.
The ‘Election Botnets’ were very different, and we’ll give you one guess at which campaign ran a more organized, more substantial botnet.
The Clinton botnet was made up of 264 bots, Trump’s botnet consisted of 680 more.
The Trump bots were far more active. They tweeted seven times to every one time a Clinton bot did the same and made far more noise around the times of political debates between the two, often declaring Trump the winner before the debate had even begun.
They were also more intelligent when it came to their use of hashtags. Most bots, and human users for that matter, use hashtags that associate their tweet with whichever side they support: #hilarysupporter, for example.
The Trump bots were using these hashtags themselves, interspersing them with anti-Clinton and pro-Trump messages and hashtags, which allowed their tweets to appear in pro-Clinton threads.
When election day finally arrived, 25% of pro-Trump twitter activity was being driven by bots.
20% of that was using both Clinton and Trump hashtags, creating discord and tension, spreading false material.
The day after the election, accounts that had been tweeting political material twenty times a day, accounts that had mimicked human activity and been thoroughly active during the day and quieter through the night, accounts that had been churning out hundreds and thousands of posts, went dead.
Someone, somewhere, had simply flicked the off-switch; the accounts just weren’t needed any more. They’d fulfilled their purpose.
So back to the question. Should we worry about bots? Well, if you value democracy, then yes.
Voters rely on information, and when they can’t rely on the information they are served, the very fabric of democracy is threatened.
Now, all this sounds very malicious, very subversive and really quite worrying.
But bots aren’t malicious in themselves, they’re simply automated social media accounts that churn out and share posts and messages.
You can’t deny social bots have real influence, and can reach a massive group of people, so, they look as if they’re a pretty good marketing tool.
You can use bots yourself, a tool that has the power to alter the course of political history, to boost your own marketing campaigns.
If that doesn’t pique your interest, then you’re not a marketer.
How Can I Use Social Bots?
Bots are useful for number of different reasons when it comes to your social media campaigns.
However, it is always prudent to keep up with social trends and ensure your usage of social bots isn’t outdated or detrimental to your marketing efforts.
That’s why we’re going to cover a couple of dos and don’ts of how you can use them in your marketing campaigns.
The first thing we’d fully-discourage is the usage of bots as a way to artificially boost your following on social media.
Issues that have arisen around incidents such as the 2016 election have led social media companies, Twitter especially, to crack down on fake accounts and ban them.
So, if your following is made up of thousands of bots and not a huge amount of real followers, once these thousands of bots start getting wiped from the site your real followers will start asking some questions.
They either question why you are losing so many followers and unfollow you as well. Or, look into why so many are unfollowing, realise you’ve had a large artificial following and lose trust in your brand.
Also, remember that social media is your way of reaching out to your audience, it’s a way of breaking through the cold, corporate front of a ‘business’ and developing human relationships with your customers.
Overuse of bots is an impersonal approach and only serves to alienate the customer when not used properly.
You shouldn’t use chatbots to handle customer enquiries. As we all know, these delicate situations require the human touch and emotional intelligence.
Chatbots can be used to supplement your customer service, by answering the simplest of questions and guiding users through the steps they need to take to find the solution they need.
However, keep their tasks simple, and be clear with the customer that they are chatting with a bot. Any pretence that the bot is a real human is only risking a loss of trust from the customer.
Chatbots are really the most useful type of bot available for you. Bots designed to share your posts can work initially but can come back to bite you if someone investigates who is actually sharing your content.
They’ll see fake profiles, and the posts themselves can often look spammy. This can cause yet more users to mistrust your brand and your marketing efforts.
We mentioned keeping chatbots for simple tasks, but this doesn’t mean you can put little effort into programming and setting them up properly.
You need to thoroughly research the type of questions your customers might ask, and ensure the chatbot will reply with personalized, useful answers.
Include links to sign-up, register, buy, etc, within your chatbots, as these have a far higher clickthrough rate than ads and offers in any other space, due to the more personal nature of a messenger box.
The benefits of this can be huge, as they can keep customers engaged with your business, offer quick solutions, make suggestions to users, guide them through the converting process and overall just help to ensure a positive customer experience.
This will help you retain and grow your customer base, drive traffic to your site and make sales based on your own suggestions to the user.
AI bots are far more useful than simple rule-based ones, as they can adapt to individual situations and have something like a real conversation with a user.
Essentially, just be open and honest with your audience, and use bots the way they were first designed to be; to provide quick, simple solutions.
Social bots are unnerving, and as we stand currently, are only going to become more influential. They’re already starting to manipulate stocks and shares, are having an influence on political events all over the world.
However, there’s no reason you can’t use them in your own marketing campaigns, and they can be a really useful tool in establishing a great user experience with your business.
Hopefully now though, you understand the power and influence of social bots. If there is a genuine takeaway lesson from this blog, we would say it is to be savvy on social media.
Don’t believe everything you read or watch, because a lot of the time, there’s a bot behind it all.
Stay in the know!
Subscribe to our blog newsletter to be the first to know when another Adversent marketing secret is out!