How Do The Social Algorithms Really Work?

Features

Instagram Algorithm

Why chasing followers is pointless

If you are on social media there is a good chance you have seen someone posting recently asking for help to reach their follower goals. There is nothing about this I disagree with. If I am scrolling on social media and it is going to take me 3 seconds to follow you and help you I will gladly do it. What concerns me is that people don’t realise how these social networks actually work. Follower numbers is something that is very low down in the list of what helps your engagement. Both Twitter and Instagram have algorithms that favour certain factors well over the actual number of followers you have, and someone with 10 followers could easily post a post that absolutely kills it compared to someone with 20k followers who posts a post that gets little engagement.

So the question is, how do the social media algorithms work?

Twitter

Remember when Twitter changed their timeline from a chronological one to one that works in a similar way to Facebook? Then made it possible for you to see tweets someone else has ‘liked’? Do you want to know the reason for this?

Twitter are WAY, WAY behind Facebook in selling advertising space (or inventory as it is called in the business). The only way they can increase the space they have to sell advertising on is by showing you more tweets. Now obviously, everyone would start going crazy if they started getting shown absolutely random tweets that have nothing to do with their interests. Thus, Twitters little tweaks were born.

When you are looking at your timeline, you are actually looking at a collection of tweets that have been gathered by Twitter and then scored by ‘relevance model’. This is the first indication follower count means nothing. If you usually tweet about beauty, it is likely that most of your followers have the similar interest of beauty. If you start going out there on a limb and tweeting about politics, it is highly likely near enough none of your followers will see it.

This is also the case with the ‘in case you missed it’ part of your feed. You may not necessarily be shown tweets from people you know, but if they are scored as relevant you will see tweets by people you do not follow.

How does Twitter rank tweets for you?

There are a number of different ways that tweets are ranked using the Twitter algorithm. Twitter now borrow quite heavily from how Facebook’s newsfeed works. There are three main factors to your tweets being ranked.

The tweet itself

The tweet itself is a big factor in how it is ranked in other peoples newsfeed. The media and language used in the tweet is one way that relevancy is determined, but also interactions. Once a tweet starts to gain interactions. So this is where the domino effect comes in. Basically you want to post out a tweet that gets as many interactions as possible in the first few minutes of it’s lifetime. These interactions then ensure the circle widens a little. Then the interactions from these users makes it widen more, and it continues. If you want a tweet to go viral you really want to be looking at getting as many RT’s and likes as possible in the first few minutes. Timing and hashtags, I have often found, are the best way to do this.

When you are A/B testing the best times to tweet, remember the above. The way your tweets appear on someones newsfeed will be completely different to the way someone else’s will. You cannot just read a generic “this is the best time to post” guide and hope it works for you.

The tweet’s author

The second thing the Twitter algorithm takes into account is the author. Randomly, the origin of your relationship is one of the factors they count. So if you both followed each other around the same time, you have more chance of seeing each others tweets in your feed. If one of you followed then the other took ages, there is less chance. Past interaction also plays a part here. If you have interacted with this persons content in the past, regardless of relevancy, you will be more likely to be shown their content now. Your ongoing connection to them is also a huge factor. This is to ensure people see stuff from family and friends and continue to use the platform. If you regularly interact with someone you will see their content all of the time.

This is why, rather than follower hunting or furrowing your brow over algorithms, the best thing really is to just make friends and interact with people.

The tweet’s you found engaging in the past, and how often you use twitter

The engaging in the past bit sounds pretty self explanatory doesn’t it? That’s because it is. Twitter will show you similar content to what you have liked and RT’s in the past.

The “how often you use Twitter’ seems a bit confusing doesn’t it?

Well let’s put it like this. Say Twitter ranks content 1 – 10, 10 being the most important and relevant to you. If you use Twitter really rarely, like once a week, then your feed is going to be filled with solid 10s from that week. If you are on Twitter a lot more often, you have more chance of those pieces of content ranked below 10 being shown to you. So engaging with people on Twitter will benefit you, but knowing who is on Twitter a lot and engaging with them more is going to be absolutely worth it’s weight in gold to you.

Instagram

Instagram is probably one of the biggest bug bears of bloggers recently. The merger with Facebook did nothing for us, and also the constantly changing goal posts seem to make it impossible to get a break. Like Twitter, they have also recently released the ‘you might like’ feature which is again down to not having enough inventory to sell to maximise profit on their ad space.

The good thing about Instagram is, although it can see confusing from the outside, it is a really simple, sequential process that you can easily learn to take advantage of. Here are the steps Instagram takes with it’s algorithm, and how to make the most of them.

Step 1

Your post is shown to a small percentage of your followers. No one seems to be able to agree on what this percentage is, but LinkedIn’s is 10% so it is likely that it will be around this figure as all of the social networks use very similar algorithms.

Step 2

Instagram measures how much engagement your post gets from this smaller audience. Again, the trick is to get as much engagement as possible. The more views, likes and comments your Instagram post gets the more likely Instagram is to bring it to a larger audience of people. It is the same as the way engagement works with Twitter, it is like a domino effect. As soon as your post starts picking up steam it just gets shown to more and more people.

Step 3

Now for the comparison stage. Instagram compares the engagement of your posts to previous posts. If it is getting more engagement, it will get shown to more people. If it gets less, it will get shown to less. This part of the sequence is why when you start to do well on Instagram it just snowballs. Posts that are getting less engagement tend to be shown further down in the timeline.

Posts that are getting a lot of engagement, especially in a short time, are shown higher up on the newsfeed and also the knock on effect means that they will be shown to other people. Posts with high engagement are the ones that get exposure on the explorer pages.

So there you have it guys. You could have millions upon millions of followers, but if you are posting content that no one engages with it will seriously get you nowhere.

Adding on to this, a really helpful Reddit user also added on that the posts are shown to a smaller audience for one hour. Instagram then works out what PERCENTAGE of your followers engaged. So not only is it silly to have ghost followers, it is actively detrimental.

Remember the value of also sharing social links on your website. Standard WordPress layouts do not offer a lot of choice when it comes to social widgets, so if you want to really build a career as an influencer you will need to go self hosted. Fast Domain are the cheapest I have found so far.

Advertisements

3 thoughts on “How Do The Social Algorithms Really Work?

  1. This was such an interesting post to read! I have to admit that I’m not all that technical with understanding the algorithms, particularly as things are changing all the time. Seeing things others have liked is a feature that really bugs me, but it’s been interesting learning about why it’s been introduced. Xx

    Tania | When Tania Talks

    Like

Love to know your thoughts guys?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s