How I started and upgraded my InstaGame with Automation

I was never a social media guy. I started my food Instagram account after getting inspired by a friend’s account, her’s was a year old and was doing good for the input that went in from her side. So I said why not!! and started the journey.
*Skip to the fourth week for automation part.

Here is my 12 weeks Journey

First Week

I posted 7 days regularly with few relevant hashtags and somehow managed to get 200ish followers, and got connected with an IG engagement group. 
By the end of the week, I learned that it’s mutual, you like mine and I like yours, and the more hashtag a post has the better it’s engagement will be.
So I started liking almost all the pics that were being shown on my feed.

Second Week

I learned that there are apps that can track un-followers and lets you unfollow them, also some of these apps let you copy followers of a user.
But they also have their own limitations and they are not automated, so one still has to put in the effort.

I was enjoying the initial rush of increased engagement and followers on my page. But that faded off by the end of the second week as I was still putting an effort on the mundane task of following and liking and not on actual content for my page.

FYI: These apps are unauthorized, meaning they’ll take your password and then store it, they don’t use OAuth for authentication. Literally, they can be termed as phishing apps, they can scam you any given day. I used them knowing all this because my account was small and not much was at stake.


Third Week — 1200+ followers

I was exploring how I can use Javascript in my browser console to automate interaction on my IG feed. 
I was able to do it with a hit or miss. It was definitely not deployable, just a proof of concept that it can be done.

Fourth Week (Amazing stuff happened)

I knew selenium’s use in automation testing and decided to leverage it and automate IG. I was familiar with python, hence went with selenium and python combo. Within a day I was able to do very basic stuff: login, scrolling my feed, going to a specific profile and following it.

Amazing Stuff ………
For some reason, I just googled “python selenium instagram”, and came across this video by Tim Großmann. I started skimming over it and realized this was the thing that I actually needed.
Got the github repo link in the description and started skimming through the code.
I started tinkering with the library and wrote my own scripts according to it.

Fifth Week

Playing with the InstaPy library and AWS server….

Sixth Week

I streamlined my scripts:
12:00 am-6:00 am — unfollowing non-followers
6:00 am-12:00 am — following users of specific users and liking 5 pics of people whom I followed.

Problem: Both my following and followers count were increasing at the same rate and I was following many bots and fake accounts along the way.

Seventh, Eighth and Ninth Week

Left the script running as it is, concentrated on the content.

Tenth, Eleventh, and Twelfth Week — 2500+ followers

I decided to unfollow people with Following > Followers count, almost halved my following count.

following chart

I also became part of few engagement groups that use different hashtags every day. So everyone will interact with the posts on that hashtag, and I automated my interaction :p. Now, I was getting 150–200 likes on every post.

To update hashtag every day on the server I used telegram bot and a modified version of a simple echo server.

Current State

As of now I follow only those users with Following > followers(ensuring they don't unfollow back) and unfollow them after 24 hours.
I also keep track of my unfollowers and unfollow them as well, baring a few big accounts. This keeps check on my following count and at the same time increases my followers.

Engagement group also help me maintain my engagement rate.

Here are my social blade stats.

By no means these are amazing stats, they are decent, I have seen accounts outperform these numbers but the admin of those accounts put in the actual effort of engaging. 
Amazing thing about these stats is automation and consistency.

Concluding this with the shoutout to Tim Großmann TimG for the beautiful library.