Below is the story of how Instagram automation can change your profile forever. Note that he started with next to zero followers and almost doubled his growth every single month.
This case study is in his own words, so we'll let Tim take over from here.
The first month (357 -> 757 followers)
I started doing some simple data analysis (including a linear regression to predict the next day’s growth) to find the best tags and parameters.
Using this data to run a simple linear regression, I could pretty accurately predict how many followers I would reach 9 days later. But on the 10th day, there was quite a big gap again.
Predicted | Actual
762 | 766
775 | 767
788 | 773
801 | 805
814 | 827
827 | 836
840 | 856
853 | 879
866 | 874
879 | 904
I did this so I could later check back to see whether the growth was exponential or just linear.
In the first month, about 13 new accounts followed me every day. What’s pretty interesting here is that on some days, I even lost some followers. A lot of them were accounts like “MillionaireMentor” and some other accounts with millions of followers.
When looking at the daily growth even closer by looking at each “run” of the script separately, we really can see that Instagram really has a lot of bot-driven accounts.
Note: 0 new Followers doesn’t mean there were no new… The unfollowing ones just canceled out the new ones.
If we take a look at the tall negative bars, we can see that over a period of 8 hours, I lost 8 followers. I only netted 3 new followers (the mean number for each session) 11 people unfollowed me.
The above picture shows the standard growth 10min after the script ran and 25min before the next session started. This is how it looked all the time. Even on the days that I lost followers.
The idea behind the tactic of following and unfollowing came from something my friend said to me: “I almost feel like I owe you something for following me.”
Most of these followers seem to be real. Not like the followers you buy that are just empty Instagram accounts with no activity at all.
As you can see in the image below, the newly won followers are actively contributing to my content. They aren’t just inactive accounts.
Compared to the average monthly growth for an Instagram user (in 2016) 16%, I reached ~112% growth in one month (growing from 357 to 757).
The second month (757 -> 1472 followers)
The second month went even better. I managed another 94% growth.
I started looking for the right tags to reach enough people. This was extremely important, since the more posts there are, the more people use that tag.
You can see tags and their number of posts when you’re about to post.
With these 3 tags, I already reach a large base of potential followers. So here are my tags in a word cloud, where the size represents the amount of usage:
To analyze my used likes, comments and tags per post, I wrote a small script which extracts all this information from the given profile and saves it as JSON for easy processing.
Tip: it’s best to use positive tags, tags which will be associated with positive feelings and situations.
If we take a closer look at the graph above, we see that for a post with the tag “tasty” I get an average of 390 likes.
I also recommend sticking with 2 or 3 categories, rather than posting images of everything.
I concentrated on 2 big categories and 1–2 smaller ones. The two big categories were “vegan food” and “vacation / goodlife”. The smaller category was “nature” (this is the nature of my home country, Germany).
To really see the impact of the script on not just my number of followers, but also the real activity of that new followers, I decided to get the likes and comments from each post. Plotting these, I got the following graphs:
In the graphs above, I plotted a vertical line for when I started running the script on a server. If you compare the complete red line with the dotted red line and the numbers in the legend, you can see that the numbers more than doubled.
Then I added the follow/unfollow functionality. I was pretty surprised by how much of an impact this had.
Here’s a screenshot I took shortly after waking up and not checking Instagram for ~7 hours:
If we compare the ~23 new followers per day in the second month with the ~13 new followers per day in the first month, we can get definitely make the guess, that the more followers you have, the faster you attract even more followers.
I was even more impressed at how strong the impact of the follow/unfollow tactic was. I grew the new followers pay day from ~23 to ~32 by just following some of the accounts I liked the posts of, and then unfollowing them again on the next run.
Based on my estimations, an account that already has ~5k followers could gain 70 to 100 new followers each day with this script.
The third month (1472 -> 2084 followers)
“Why do you keep following and unfollowing me all the time?”
That was a problem I haven’t thought of before but was pretty obvious. I fixed that problem pretty quickly because I thought that it could be an indication of automated accounts.
My solution to keep a dictionary of accounts I had already followed, and I can define a maximum number of times I want each account to be followed.
Since I hit 2,000 followers this month, I hosted a small giveaway. This not only gave me a little follower boost, but it also made my Instagram account seem more legit.
Giveaway of two coconut bowls
When thinking about it now, it probably would’ve even be better to indicate the giveaway in the picture directly instead of just writing it in the description.
If we take a look at the graph above, we can see that the mean number of new followers dropped from ~32 to ~20. On days where my script had problems, I got fewer followers. Some of the days even were really good with ~40 new followers per day.
If we compare the number of new followers before and after the start of the giveaway, we can see that it gave a little boost with ~28 new followers per day.
Upon posting a new image or video, by now, I normally reach ~100 Likes in the first hour.
The fourth month (2,084 -> 2,706 followers)
In the fourth month, the growth, sadly, again was just ~620 new followers.
After 4 months of running the script, this is how my average post looks like:
Here, again, we can really see what an impact the script had. Even on the older posts, which have been there, before the script started.
As of writing this, I have ~2,800 followers!
So as you can see, genuine growth can happen on autopilot. Tim went from a few followers to over 2,700 in just a few months. And these are genuine real followers (he did tests) who are in the target market he was focusing on (#vegan).