I have my own software that does all the heavy lifting for me, but I’d like to pair down followers on my accounts ie remove fake followers/bot accounts etc.
I store and update daily simple metric for each account that follows one of my accounts (albeit basic information) - ie number of posts, followers, following and date of last post.
Is it as simple as saying unfollow all where postcount is less than 0 and following >(2 x followers) sort of thing?
I dont have an algorithm per se, but I do use Jarvee to eliminate accounts that follow me that are what I call “ghost accounts.” I first get a list of my followers. Then, I block all accounts that do not have a profile image and all accounts that have less than 10 posts. Then sort it their following count (the accounts they follow) from greatest to smallest. Then I filter out the good accounts that has more than 55 posts (most bots do less than 50, not all). I tackle the remaining bad list from top down, almost anything above 5K is just a bad account to have following you. I also have started to make historical following pulls, and I will compare a 6months ago to my current list. If I have a follower that has not liked a single post in 6 months, I block them too. Best way for me to keep my engagement up. Have not seen any drops in likes, in fact it really helps my account health.
This is the ONLY thing that is helping your account.
Any speculation that a follower with above 5K followings is bad or an account with less than 10 posts means inactive is just a waste of your time.
There are a ton of lurkers that like and comment on a zillion accounts and have no profile picture and little to no posts. On the flip side, IG botting 101 is to profile the account, if you see where I m going with this.
Engagement is binary - a person engages or doesn’t.
On that note, an account with 6999 follows that has 0 posts and no profile picture that engagees with your content regularly is a lot more valuable than a nicely profiled account with a ton of posts that ignores your content on the daily.
If you only want people that are engaging with your posts, extract the list of users that have liked & commented on your posts and compare to total follower list. Although you’d have to do a lot of blocking then lol.
Also, often people just like to view content and don’t care about liking or engaging. Just watchers. It won’t do much for your “engagement ratio,” but if they like your stuff, who knows, maybe one day a friend of theirs who likes the same stuff will find them through recommended users and give you a follow. And real engagement.
Blocking is only useful for obviously bad accounts, bad being relative and arguable, beyond the clear cut spam accounts.
Though, sure, I do block lots of “bad” looking accounts with 6000+ followings that never like or comment, and don’t have any photos posted, because I know they’re just automating with f/uf tactics.
These are dead weight accounts, they do affect your impression ability. It’s also obvious in the like data too. 99% of these accounts have given 0 likes after 6 months. So its pretty proactive to get rid of them. Especially if they are an active bot, that is still posting pictures, you will get post impression on their account, and no likes. Terrible for engagement.
Also, zero profile picture is a huge red flag for IG if you have too many of these following your account. I worked closely with DeepSocial who had a license with IG to create a product to help social account analysis. Zero profile picture was one the biggest factors they focused on. Take it as you wish, my guess is the data for likes will show you exactly like I said with the 5K+ following count accounts, 99% are not liking your posts. My actions are preventative, where I am not waiting 6 months to find out that account is complete dead weight.
I’m not trying to be an ass for the heck of it, just want to clarify what is the data talking and what is your guess.
Also, what about say 4k to 5k followings? Do you have any data on those?
====================================
In general, the whole notion of a “fake” follower is lost on me.
Who decides who is fake and who isn’t?
Is a “fake” / bot account that engages with your posts worse than a legit IG personal account that does not?
Not to mention that any respecting bot account owner would profile their account and put some content on it, to avoid looking fake. I would even go to argue that more non-profiled accounts are actually legit ones.
Looking at my own personal feed of 178 followers who are actually people I know in real life, 20% of them have 0 posted content, but are for a fact quite active. Some of them have no profile picture. Back in the old old days, when IG was just a photo app, people were having zero data on their profiles as they used the app for cool filters, not for the social media aspect. Some of these people still remain.
My whole point is that all of these are speculations. We don’t know the full picture so blocking people on criteria other than engage / does not engage is potentially wiping out people who engage with your content. Do you get a reward or a penalty for that? Only IG knows.
My whole point is when you are scaling an account these things do matter. Everything I have done is off data, I pull likes for a whole year. I look at all the factors that lead to preventative action and the things I outlined are the most common offenders. If you want to keep the small exceptions, be my guest. But a huge portion of followers that follow that fall in those categories are not engaging. So don’t throw your personal account’s abnormalities and tell me I am guessing, when all I use is data to make decisions.
I’m sorry but this is a stretch and I don’t base the most important metric on IG “engagement” on the off chance that a red-flag account is going to have a friend follow me. If anything, there is a remote chance that follower of an account that doesn’t like my content and follows thousands will even follow me in the first place. I don’t live by outliers, I look at the data and those kinds of accounts are more risk than any single thing a small few might offer.
