So now we have a much lower follow limit since 6/4, we have to get even better at increasing the follow back rate and not wasting as much time following the wrong people.
A little background, I have been running my homebrew F/UF program for a almost a year. I started with using Instapy but just ended up subclassing and kept on adding my own twist. Now my program is attached to a full SQL Server database where I keep track of all my activities so I can do analytics on the results of the actions and constantly improve my program accordingly.
My targets are people who frequently comment and like similar accounts which I keep track in my database but one thing I could never figure out how to do is how to avoid the engagement groupers and their non genuine likes/comments. I feel like if I can somehow identify them, I can increase my follow back rate a lot. Anybody has any ideas or suggestions? I have been scratching my head for a long time on it and the only thing I can come up with is if a similar account’s likers/commenters give me low follow back rate, it’s possible that’s a engagement grouper so I stop following the rest likers/commenters under those similar accounts. But it’s kind of just guessing and I wouldn’t know until I waste 50-100 of my 6000 on each of them. There must be a more efficient way, I just need to figure out what. Maybe more heads are better than one.
Hi, im using a similar strategy and have a similar issue, I import scraped data into SQL server and then analyse with various queries. My f/uf is all manual though and very time consuming - have you automated this in a human way with all the logging?
Watching the thread closely as ive also been scratching my head trying to think of how to avoid the engagement group users
I automated with my customized subclass of Instapy. Human way is kind of subjective. Insatpy uses chromedriver and selenium so it’s like the program is clicking on buttons on IG website. My follow and unfollow is done in (now) 200 batches with random 45-75 second in between actions then 2 minute +/- breaks between random 8-12 actions. I have been adding my own story views, comment likes, post likes functions in the past month (instapy had some of it already but either they didn’t work too well or I prefer my own implantation, instapy definitely doesn’t log everything in a useful way which I think is way more important than automation if you want to constantly improve your process).
If you have the potential followers in the the database, you may as well record all your F/UF, story view, comment like, post like actions so you write queries to match your new followers to see your rate of return for those actions.