Around 6,000 tweets are tweeted per second. That equates to approximately 500 million tweets being sent per day! With such high traffic on the platform, how does Twitter decipher what tweets you should be seeing?
Twitter’s timeline can be broken into three parts – ranked tweets, ‘in case you missed it’ and remaining tweets. The first two sections feature only a limited number of tweets with the bulk of what you see being displayed in the remaining tweets section. Twitter’s method behind this format is to cater for various user timeframes. It aims to show you the most important and interesting tweets in a short and concise format, while also catering to those who want to spend more time on Twitter and not miss any updates.
Machine Learning Algorithms (MLAs) are common among popular social media platforms such as Instagram and Facebook. Similar to these platforms, Twitter uses an MLA method, but only for the ranked tweets and the ‘in case you missed it’ section of the timeline.
An MLA aims to show you the content that will be most interesting to you. It works to eliminate what you will find irrelevant to keep you scrolling and engaging with the platform. On Twitter, this type of algorithm considers factors both about you and the tweet itself. Twitter’s algorithm aims to consider a range of factors which contribute to determining how relevant a tweet is to you. The algorithm ultimately assigns each tweet a relevancy score by considering the following factors:
- The tweet itself – recency, the presence of media and total interactions
- The tweet’s author – the strength of your connection to them based on your previous interactions
- You – your previous interactions with the platform and the content you regularly search for / interact with
Now that we’ve cleared up how Twitter’s MLA broadly works, let’s look at it specific to the three parts of your timeline.
The first section you see when you open your timeline is the tweets that are categorised by their MLA relevancy score. The higher a tweet’s relevancy is, the more likely it will be displayed in this section.
This initial section focuses on your interactions with the platform. The MLA monitors what content is common amongst the tweets that you click on, retweet, comment on and like. It also monitors who you are interacting with through the platform. All this information is collated and used to determine what tweets reflect what you are interested in the most and thus, used to determine what you should see in this section.
‘In case you missed it’
Never seen this section before? Well, that’s because it is only shown to people who have been inactive on Twitter for a lengthy period of time. The aim of this section is to show you a few important tweets that may be very relevant to you but classified as ‘too old’ to be in the first section of your timeline. The MLA analyses content in a similar matter to the previous section but works with a pool of tweets that exclude tweets that will be featured in the first section as well as ones that have an older timestamp of a few hours, days or even weeks.
This section is surprisingly different from the previous two in terms of how it orders the tweets from those you follow. To order the tweets that have not been featured in the ranked tweets or ‘in case you missed it’ sections, Twitter uses a reverse chronological ordering method; the most recent at the top and then descending into older content based on the time that the tweet was posted.
On the surface, your timeline looks relatively straightforward, but behind the scenes, Twitter’s algorithm works constantly and almost instantly to provide you with the most relevant and up to date content.
Which MLA based platform do you prefer – Twitter, Instagram or Facebook?