3 Reasons The Social Media Algorithm Is Not Working
Am I the only victim?
Is no one else experiencing this?
Of course, you are.
The social media algorithm is broken.
Now, I am no tech expert. I have a cursory knowledge of how the algorithm works.
Most of it is from personal experience and digging around technology sites.
I can only discuss how I perceive the hitches in the algorithm. I can also be way off with my assumptions.
The Old Attachments
Maybe the premise of the entire post itself is wrong.
It could be all these apps are at their best behavior. The development could be going exactly how the owner wants.
After all, when it comes to apps like Instagram, the executives do not depend on you anymore.
Even with the dwindling user experience, we feel we have no choice but to stay on the app.
Our lives are ridiculously entangled with the internet.
When was the last time you intentionally tapped the Facebook icon?
Yet, the reminders of some distant Facebook friend’s birthday won’t leave your notification tab. So, you tap it anyway and send a quick wish.
Congrats, you are now a part of the 2.9 billion monthly Facebook users statistics. Never mind that you open the app for exactly 2 minutes every day.
Our Dependency On Social Media
If you own a business or want to branch out from the 9 to 5 routine, you are stuck on social media.
A brick-and-mortar shop alone is not enough. Moreso, if you are in the service industry.
You need social media to grow your business. If not your business, then your personal brand.
Or sometimes, you just want to be an influencer.
There is the potential for good money there, even if most people never end up achieving it.
You might get free deals and sponsorships from time to time. Again, rare occurrences depend on your effort and luck.
The bottom line, you can not let go of these platforms. Whether it is YouTube, Instagram, or LinkedIn- you want to give online fame a chance.
And why would you not?
Unless you are born into generational wealth, social currency matters.
How Influencers Depend On Social Media
Take LinkedIn, for example. It used to be a job searching site for professionals to connect to each other.
While it was always intended to be a social site, the intensity of people’s interaction on the platform has changed.
LinkedIn is cluttered with influencers, and you can divide them into two groups.
- Freelancers who want inbound traffic.
- Employees who want to keep their options open.
Visibility allows freelancers to reach their ideal clients.
Even those who do not get big on the platform accumulate a small list of followers.
There is a benefit to reminding people you exist.
The second set of people do it as a failsafe. Suppose they quit their job suddenly or get fired, they will have their followers to fall back on.
They will have a steady source of income on the side or connections who can get them a job quickly.
What About The Regular Social Media Users?
There are people who are on social media because it is the norm.
They want to get updates on their friends’ lives or see funny tweets.
Some feel irrational envy over someone else’s aesthetic Instagram lifestyle.
Others are busy posting about the most depressing times of their life for laughs on Twitter.
In the end, we are all victims of the algorithm.
Now, I am not one of those people who advocates how we should get off the internet.
Especially when your complaints are posted on the same platforms, you are encouraging people to leave.
To have easy access to the internet is a privilege in some parts of the world.
So, it feels redundant to complain about any of it.
I can curse social media all I want.
I can also acknowledge how I have witnessed it change the lives of dirt-poor people in my country.
The point has always been to strike a balance.
Don’t go filming someone’s worst moments for views on social media, and you are good.
As long as your virtual world does not take priority over the real world, you can find some sort of stability.
In the end, developers do not care all that about mental health. All platforms have one thing in common.
They want you to stay on the app as long as possible.
How Social Media Predicts You
Ideally, social sites want you to keep doom-scrolling until you come across an entertaining advertisement.
It will be something you mentioned out loud you need or was a passing thought.
Not a speck of coincidence there.
Is the algorithm capable of reading your mind, then?
No, but it is capable of predicting human behavior.
Apps collect your data so heavily that they can form a pattern of your internet activities and, by extension, your life.
Odd thing, I never minded it much.
I am good at not making hasty purchases.
If the algorithm can show me what I need when I am about to need it, what is the harm?
If the content aligns with my interests, why not let it?
I was fine with it when it was working like a well-oiled machine.
It did not cause paranoia as it does with some people.
In part because I am well-versed in social media detoxing. It is not a conscious decision I have to make.
It is something my brain forces me to do.
The introverted side of me extends to the virtual world. I lurk rather than post or interact.
I can distance myself when I need to. (I am more likely to lose time obsessing over a tv series than general social media.)
But the fine line of balance I rely on the algorithm to maintain is not there anymore.
When Did The Social Media Algorithm Start Declining?
I can trace the timeline of the glitch back to the coronavirus pandemic. People felt the importance of the internet more than ever.
When people were stuck inside four walls, there were not many ways to stay in contact with the outside world.
Some had the fortune or misfortune of living with their family or roommates.
Others were alone and craved human contact by any means.
People lost jobs and looked for alternatives to sustain themselves. Some had enough savings to spend time fulfilling their creative ambitions.
Short-form videos turned out to be the general public’s preferred way of expressing themselves. This brings us to the meat of today’s post.
The three factors which changed the algorithm:
- Short-form videos
- SEO
- AI
The Three Main Parts of The Algorithm
1. Short Form Videos
You can blame the short-form video revolution on TikTok.
It is like the modern-day Vine.
Whereas Vine showed us what six-second clips could do, TikTok took it a step further with fifteen seconds.
TikTok’s popularity took off before the Covid-19 pandemic. However, it was the teenager’s app.
Now, we can not get solid statistics, but 38.5% of TikTok users are within the 18 to 24 age range globally.
According to Pew Research Center, 67% of US teenagers are on the app.
So, while it is hard to get confirmation, it is safe to assume the biggest demographic was and is teenagers.
The app was an instant hit when it came out in 2016. It used to be people lip-syncing to different songs.
Over the years, the content became more diverse as the time limit went bumped to the present ninety minutes.
People make whole films on TikTok.
The app was on its way to Instagram and Facebook status by 2019. A feat that was accelerated due to the first lockdown of March 2020.
Stuck at home with limited stimulation, social media did become our only outlet of connection to the outside.
Naturally, people documented more of themselves.
When the surge of videos flooded TikTok, the developers naturally saw an opportunity for more revenue.
The advertisers flood wherever the audience is.
TikTok gave creators the opportunity to get paid for their content.
Combining free sponsorships and brand deals, the forecast for influencer market trajectory is $249.65 billion by 2033 for a reason.
Naturally, the algorithm got a round of updates. Keeping people on the platform- addicted- was always the goal, and it was reinforced.
I will say TikTok is one of the few social sites which was not completely ruined due to constant updates.
TikTok’s algorithm is pretty smooth running still.
Mostly because it remains a short-form video platform and does not claim to be anything else.
We can not say the same for Instagram, Facebook, YouTube, or Twitter.
Once they got wind of TikTok’s strategy, Meta (Facebook and Instagram) had to capitalize on it.
The fact TikTok got banned in large markets like India helped Instagram push short-form videos or reels to a specific audience and get the attention of the pre-existing user base.
YouTube followed soon with shorts, and Twitter joined the game last.
Although, Twitter has the biggest issue promoting short-form videos due to its user age demographic and the general tone of content on the platform.
Now, we know the USPs (or old USPs) of these social platforms.
Instagram is for pictures; Facebook is for life updates. YouTube is for vlogs, and Twitter is where you vent deeply personal stuff to complete strangers (and lots of fandom stuff).
What happens when social platforms with completely different content types force everyone to conform to one medium of interaction?
People do not leave, but they can not stay either. Some adapt and use it to their advantage.
Others fold under the pressure and hop from one platform to another- unsatisfied with them all.
The algorithm takes the biggest hit.
It has to account for all the familiar content mediums and introduce new ones.
As a result, you can not see your best friend’s photos on your Instagram Homepage anymore.
What you do find are home renovation pics from that one account you liked two days ago.
Sprinkle that with lots of irrelevant advertisements, and you do not feel like scrolling your home page.
So, you move over to reels, where the content is interesting if from strangers.
After all, it took into account everything you have liked, saved, and shared on these platforms and suggested accordingly.
Except, you can like one reel relating to someone’s existential crisis, and it is the only thing you will see as you swipe.
Because the algorithm has deemed it as your primary personality trait.
It is worse with YouTube, which is still in the middle of hard-launching shorts.
The algorithm appears to have an idea of the content a user likes but no semantic recognition.
I have no idea why the algorithm will suggest massively conservative, deeply misogynistic content to me otherwise.
2. SEO
Social media sites have started using keywords as signals to determine the relevance of the content to a user.
This works on three levels. The search engine crawlers are smart enough to read texts on images and videos along with captions.
However, the information verification process is not as effective as it could be. For search engines like Google, the crawlers are more equipped, as there are metrics the content has to meet.
Things like backlinks from authority websites and citations can quickly get content to the first page.
Even so, Google still comes up with faulty answers in the “People also ask” corner. Competent search engine optimization experts know all the metrics and how to manipulate them to push content that may not be the absolute truth.
It has led people to tack on a “Reddit” at the end of their searches. After all, as divisive as the opinions may be on the forum, it usually comes from real people with lived experiences.
This hole in the system is amplified on social sites. The virality factor of the content appears to trump any need to follow SEO rules.
They push keyword-heavy content to your feed, making assumptions from your previous likes. If you watch or interact with the content, you will encounter more of the same.
Does not matter how inaccurate it is.
You get a repeat of the same keyword and keyword over and over again. Often in a way that makes you suspect the site is going through a breakdown.
I liked content from the Northeast Indians ones, and the only thing I saw for the next 20 reels were people from the same region. Clearly, it picked up the location part of the keyword but failed to account for other components of my likes and dislikes.
Before, you could at least tap on the “Not Interested” option, and it would work. Now, sites like Instagram can not seem to get the obvious signal.
There is an obvious lack of understanding that liking a part of something does not mean you have to like the entire niche.
No social media site appears to get when I like BTS content, I only want to see them. I do not need to see other K-Pop idols or Korean dramas or actors.
I have tried signaling it to these sites by blocking and clicking “Not Interested,” and it never works.
3. Paid Verification Badges
The verification badge is the latest ingredient in the melting pot of disaster decisions.
The blue tick mark was something reserved for political parties, celebrities, businesses, government organizations, and other public figures.
The badge had to be earned through achievements in your field. Published writers had as much chance of receiving the badge as the actor of a popular TV show.
It is a symbol of your contribution to society.
All of it went down the hill when Twitter introduced the paid badge option. Anyone can get verified as long as they pay a subscription cost of $8.
There are two benefits to the badge. One is the illusion of importance it adds to your account from a public viewpoint. Second, is the preference your content receives from Twitter for paying.
For a while, Twitter even aimed to remove legacy verification in an attempt to force everyone to pay.
It did not work for long, as massively established public figures refused to take the deal.
As such, anyone with over 1 million followers is now verified, with or without a Twitter Blue subscription.
However, it still means many lost their earned badge and their platform.
Now, Meta is rolling out the same feature for Instagram and Facebook. You are set to increase your reach if you sign up.
The good thing is Meta does not plan to interfere with the accounts of previously verified users. The bad news is that it would be ten times harder to receive the badge without paying.
It is also another factor that will spectacularly mess up the algorithm. As it already has with Twitter. For sites that claim to focus on content-based reach, I have seen absolutely dull tweets from verified accounts show up on my timeline.
What can developers do to make the algorithm more bearable?
While I would prefer to go back to the good old times, I doubt it is possible. These social platforms followed the age-old business model and won.
You reel in people with free stuff, and when they are addicted, you find ways to make them pay. Right now, users are paying both in the form of advertisement revenue and subscriptions.
What they can do is make amends and bring back some of the familiarity which users love.
Take a look at the complaints users have with the social media platforms. If they can remove these features or notch it down to something agreeable, plenty of customers would be satisfied.
- Bombardment of the same account on your For You Page if you so much as click on their post.
- The “Discover More” section under every tweet only servers to irritate users.
- Dividing the timeline into “For you” and “Following” sections. Users want their feed to have a mix of interesting content and people they are willingly following. They do not want to make a choice between one or the other.
Instagram and Facebook
- Repeatedly pushing content you have clicked “Not Interested” on.
- Constant odd changes to the layout and tools.
- Showing everyone but people you follow on the feed.
- Display of content based on relevance over chronology.
- Limiting the circulation of feed posts to a ridiculous level. Especially since reels get triple the time.
YouTube
- Recommending content that has nothing to do with the keyword searched.
- Repeatedly showing the same type of content despite clicking “Not Interested.”
- The lack of a block button. While the “Don’t recommend channel” option exists, YouTube seems to think of it more as a suggestion.
Final Thoughts
The concept of a content-based algorithm fails to take into account our real-life relationships with people. Yes, I might not have any interest in this TV show my friend keeps raving about, but if she posts about it? I want to like her post. It is an act of support and a way to know what is going on with their life.
I do not mind boosting content from total strangers, but I also want to be able to do the same for my friends. So, more than anything, a true mix of content and following based timeline would be great.
