How To Do a Proper Test (Follow Block or Any Other)

When a big issue comes up, the forum gets filled with strategies and suggestions on how to solve it.

Some are better than others, but I believe that testing for yourself is the only way to see what works and what doesn’t.

But…

You need to do a proper A/B split test.

What does that mean?

Your accounts (batches), need to have:

  • IDENTICAL software settings, to the comma

  • IDENTICAL proxies

  • IDENTICAL phone number verification method

  • As close as possible or identical account history - how there were created, when, what they did to warm up, etc.

Then, you make a SINGLE change to one of the batches, leaving everything else the same.

Then, you give it at least 48 hrs to actually run.

Then, you come back and compare the results to see if there is a difference.

IMPORTANT: the bigger the batch, the more conclusive the results (of course). An account doing A or B does not really prove anything, so try to test at least 50, ideally 100 or more accounts.

IMPORTANT #2: be as objective as possible. Testing means looking at what is going on and drawing a conclusion. It does NOT mean doing a test so you can prove a biased theory. Avoid looking only at the results that best fit your theory as that is pure BS.

What is a BAD test?

The shortest version: testing something on accounts or batches that have little in common.

You see it done on the forum all the time: “I tested X vs Y and for sure, 1000% X works”. Then you dig a bit deeper and you see that one batch has mobile proxies, the other one has DC proxies. Or one has been purchased online 7 years ago and the other one is brand new. Or a client account vs a slave account.

When the batches are different, you cannot test a single option as there are too many variables. Ex: you might have reset the device IDs and that got you out of a block, but the reality is that you gave one account some rest and you didn’t the other and the resting is what got your out, not the reset.


At the end of the day:

  • There is no magic solution, so you need to find what works for you

  • Everyone has different situations so what worked for everyone, might not work for you

  • NOONE cares more about your accounts and business than you so don’t just sit and wait for the magic pill to land on your lap and solve your problems.

  • Test every theory for yourself.

Good luck and if you feel anything else should be added to the list above, let me know and I will edit it in.

23 Likes

Yes, nice explanation, I’m glad you posted this, because I get a ton of messages of people asking for help and the solution is usual just test stuff. I will start sending them this topic. In my opinion this is the most important part in Internet Marketing. Do more testing people! It will always be positive.

1 Like

Thing is we have no idea whats going to happen after the service shortages are resolved and IG receives an insane amount of reports from normal users that got the ban hammer as well. So testing right now, would not be wise unless you are losing a lot of money on the back end.

My main goal for posting this is that people seem to be blindly following X and Y theory with varied results Then they come to post their biased findings or failures, fueling another round of people blindly following X and Y.

Then people get inevitable banned or slapped by IG or their platform of choice and they start blaming factors that have zero to do with it.

Your accounts, your responsibility… so test, test, test.

There is no bad time to test things. You cannot just wait for IG to solve what may or may not be a bug. And even if it is bug, the landscape has for sure changed, so better figure out your next move.

Some people are still waiting for the likes to come back on DC proxies :wink:

3 Likes

I am definitely into the taking things on your own hands mentality, don’t get me wrong.

It just feels like the dust hasn’t settled on this update yet. If they are encountering shortages right now their backend is overwhelmed and it wouldn’t surprise me if more stuff are in the pipeline.

100% agreed.

My post is more aimed at people who are quoting some theories as the panacea for the current situation… without having properly tested it or tested it all.

Example: this morning I received a DM from a person who I’m not going to name literally asking me if I support one of the suggested theories as the wanted to send an update to his clients.

“Did you test it for yourself?” was my reply.

His was: “No, but it seems to work for other people so probably it will work for me”.

1 Like

Hahaha, we need to test by ourself.
because we have a lot of different constrains.
We cant just copy without finding the solution.
Thanks i got the big picture.

Perfectly summarized. Thank you!

thank you so much for sharing this informative method

there might be same actions limits ,but as you said in most cases ↓

different account creation methods
different settings
different pv methods
…a lot more

that’s a Fact