Why social media algorithms and authors don’t mix
I’ve written about algorithms before. But with all the changes on social media, it seemed like a good time to revisit this most frustrating of book marketing topics.
People are predictable. You might think that’s not true. After all, people do weird things all the time. But they’re only weird to you. They are not weird to the person doing them, at that time, in that place. It’s all about variables.
On the whole, if you know all the variables, people are predictable. That’s why algorithms are used so extensively in social media.
The trouble with variables
The thing is we rarely know all the variables. That’s why when it comes to social media, algorithms are the quickest way to get in a rut, and the hardest rut to get out of.
When it’s noisy or I’m tired, I want funny memes and inspirational quotes. The rest of the time, I want in-depth articles and thought-provoking blogs. But social media algorithms don’t know that. They keep ordering my feed the same way.
It’s the same for readers. You can see what a book link is about without clicking it. So you may scroll past – but that doesn’t mean you will not at any point in the future want to see another one. Like when there is a new book available or you want to buy a gift.
The trouble with assumptions
Scrolling past a post with a link does not mean you don’t want to see any posts with links from that author. You may not want to buy a book right now but you may want to read their blog or a free short story. Or vote in a cover contest. Or just wait and see what the next post is about.
People don’t usually make decisions as quickly as software does. There are several variables social media algorithms use to make decisions. And there are problems with all of them.
The trouble with hashtags
Hashtags are a minefield because they are collectively managed by users. Anyone can add any hashtag to their post. The biggest challenge I have with what I see in my feeds is relevance. I am inspired by nature and wildlife. Which on social media is often accompanied by really broad hashtags like ‘animals’. This is presumably why my Instagram feed is currently full of guinea pig videos. Which is why I’m not spending much time on Instagram at the moment. That works for managing what I see. But what about the content I want to show?
If I am seeing guinea pigs instead of nature reserves, what are the chances readers are seeing my book posts? Very slim indeed, I’d say.
The trouble with media types
To an algorithm, an image is an image. A photo of an up-and-coming author on a research trip is the same as an image of a quote. Or a photo of a place we find inspirational. But it isn’t the same to us humans, is it?
I rarely make videos so my posts aren’t usually shown in the feeds of readers who like lots of videos. When I do post videos, they don’t tend to do well. This is because the algorithm has decided the followers who might actually want to see that type of post, don’t want to see posts from me. Which is a bit of a leap really.
This has become more of an issue lately, with other platforms trying to compete with TikTok. Videos are often prioritised over images and text based posts. On Instagram, it seems you are now more likely to see a random video from someone you don’t follow, than an important update from someone you do follow. The algorithm decides which types of media we see. Our choices mean very little.
And then there are links. I post this blog to social media. I’ve noticed that Twitter, in particular, suppresses it. They don’t want anyone clicking a link that will take them away from their social media scrolling. Even if they are planning to come right back.
The trouble with not knowing our limits
How much of our time on social media is wasted? By ads selling things we don’t need. By suggestions of accounts we don’t want to follow. By videos we don’t want to watch.
We know that once we have purchased an item, we don’t need any more. We know that one picture of a tropical island each week is enough. We know that a guinea pig is not a wild animal.
But social media algorithms don’t know that. Because they don’t know all the variables. They don’t know that you have already bought that book or pair of shoes. They don’t know how sick of irrelevant videos you are. They don’t know what the thing is the image is.
They don’t know you.
The trouble with measuring behaviour at group level
Algorithms don’t understand that some people click like, and drop comments, without a second thought. While others only click the like button twice a year, and haven’t quite worked up to leaving a comment yet. If they have read one of my posts, they will probably want to read another one.
But if they didn’t click the like button, they may not get the chance. While someone who did click like, maybe just because they enjoyed the photo or the title, now sees similar posts every week, whether they read any of them or not.
The trouble with labels
Once you’ve been given a label, it can be difficult to shake it off and prove you are something else. It’s even harder to be recognised as what you actually are – lots of things. Because that’s the point of algorithms. They simplify complex decisions.
I was slow to build up my follower base on Twitter and now the algorithm has labelled me a follower, not an influencer. Despite the fact that I have thousands of followers of my own now. That’s thousands of people who don’t actually see most of my posts.
They do see my replies to other people’s posts though. I get more responses to my blog through other bloggers’ posts than through my own. I also see more responses to other people mentioning my books, than to my own book posts.
Algorithms favour a consistent posting schedule. In December, I tend to post more on Instagram. Because I have more to post. The rest of the year, I post a couple of times a week. I’m fairly consistent. But does the Instagram algorithm know I’m consistent? And therefore a safe bet for potential followers? No, it does not. It takes a long time to catch up.
If you like a business page on Facebook, its algorithm will check the kinds of posts your friends engage with, and will suggest following that page to any of them it thinks might be interested. This is great if you want to help a small business, like an author, get more visibility. But if you want to keep your interests private, you might want to use the follow option instead.
Facebook also checks the number of other people who have liked or followed a page before deciding which pages you might like. Which is why you’re more likely to see suggestions for supermarkets than small businesses.
The trouble with responsibility
Content regulation is another issue that has affected me recently. Social media platforms have a responsibility to regulate the content posted on their sites. What they don’t have is a good plan for how to do this. There are two elements to regulation – monitoring and enforcement. As we know, monitoring is not social media companies’ strong point.
When it comes to enforcement, they have two blunt instruments at their disposal: suspension and suppression. Neither of these measures stop people inciting hate or enabling crime. They can just open multiple accounts and carry on. Both of these measures do impact valid, contributing accounts. Authors, and other business owners, can spend years building an audience to find their posts are no longer being shown to most of the people who usually read them. I am experiencing this with Facebook. I don’t know what they could have a problem with in my content but they are suppressing my posts. Or they may have taken a leaf out of Twitter’s book, and identified links as an enemy.
The unfortunate people whose accounts are hacked have to start again from nothing. Social media responsibility is an area where improvement is urgently needed.
The trouble with humans
We’re tenacious creatures, aren’t we? We may not be able to beat the algorithms but we can change our own behaviour to make them work better for us. If you would like to support an author, or any small business, there are things you can do.
Like, comment on, or share their posts. If you haven’t already, find their Facebook page and click the like button at the top. You can follow them on multiple social media platforms. That way you are bound to see something from them!
If you follow someone but haven’t seen a post from them in a while, check their profile. They have probably posted something you just haven’t been shown. Liking that post will increase the probability you will see the next one.
The trouble with change
There’s always a risk that change will work against you. Social media change is not forged by the people who use social media. It’s forged by the organisations who own the platforms. And they are trying to make money, not enrich our lives.
Change is generally slow. Which can be a good thing. We don’t get overwhelmed by change when it happens slowly. But it’s also frustrating when we’re waiting for it to happen. There are some good changes happening now which may improve things in the future.
For example, there are types of artificial intelligence that can read images. They know the difference between a quote, a book, and a person. They may one day be to tell the difference between an animal in the wild and someone’s pet.
There are also tools for analysing ‘big data’. That means the massive volumes of data available to large organisations, like social media platforms. Everything you do (or don’t do) on their app, is collected by them. But wrangling that data into usable intelligence is a real challenge. Big data analysis tools can help them understand the numbers of people skipping the posts in their feeds, and spending time going to profiles to see posts from people they know instead.
I’m hoping these developments will start to have an impact soon. Because I don’t know how many more guinea pigs I can cope with!
Which of these issues have you been impacted by? And what change are you most hoping to see in the future?
Enjoyed this post?
You might also like: Worse Than Being Talked About.