Arch Ramia
925 posts
@RoseRomia
lecturer . Researcher 👩🏻🏫. PhD candidate URBAN RENEWA🙏L PHD #phdfamily #phdlife
How universities can control what you see on social media: The use of bots is commonly known, but that is just one way that universities can control what you see. Another, much more effective, method is to have many accounts operated by real people. Using these accounts allows universities to control what you see in four easy and different ways. To understand how, we first need to understand how social media algorithms work. When you post something, all the big social media platforms then show your post to a small number of your followers. If these followers interact with your post, for example, like it or repost, the algorithm registers this as a high value post and shows it to more people. Initially, it is to more of your followers, but as time goes on and your post continues to get a high level of engagement, it gets shown to more and more accounts that don’t follow you. That is how you break out and grow rapidly. The four distinct ways accounts controlled by universities can control what you see are: 1.Engage with a post they want you to see, from a university controlled account. That can be through liking it, reposting it, commenting on it, expanding it to read the thread, or even reading people’s replies. That tells the algorithm that this post is high value and to show it to more people. 2.Use these artificial accounts to pump out many posts and flood the space. That greatly reduces the chances of you seeing posts from non-university accounts. 3.Follow other accounts and not interact with their posts. If you have 100 accounts see your post and 90 are controlled by universities. If these 90 don’t interact with the post they don’t approve, that tells the algorithm that this post is low value and it will stop being shown to other accounts. Hence, the remaining 10 accounts, that are real, need to engage with your post even more to overcompensate for that damage the 90 artificial accounts are doing. 4.Only interacting with posts from non-friendly accounts that they approve of. That boosts the posts that are good for them because the algorithm registers that these posts are higher value and shows them to more accounts at the expense of other posts. With hundreds of thousands of accounts in the academic space, it seems hard to think that universities can have that many artificial accounts - enough to control what you see. But is it? Let’s do the numbers. Let’s say it takes 15 minutes per day to administrate one artificial account. In other words, a real person is using that account and engaging with certain posts and specifically not engaging with other posts. A full time worker does 7 hours of work per day. That means this person can control 7x4 accounts each day = 28. Taking the least economical case possible where the university pays the person a salary on par with Switzerland (one of the highest income countries in the world). It costs the university about $80,000 per year to employ this person. If the university spends $2 million per year, it can employ 25 people, giving it control over 700 artificial accounts. That may not seem like a lot of accounts…until you realize that there are about 20,000 universities around the world. And even if just a few hundred universities do this, that results in over 250,000 artificial accounts! Most universities have the resources and money to do this, by the way. That is with the worst finances possible. If they were to outsource this work, which is standard practice, the cost to control 250,000 artificial accounts drops to less than $50,000 per year for each university (assuming each worker gets paid $1,500 per year including benefits and taxes, etc., which is a standard wage for this type of outsourced work, plus $10,000 for various admin costs.) $2 million…👇 #PhDVoice @PostdocVoice
