My Journey - Building a Streaming Bot Empire

Disclaimer: I won't be naming the Streaming platform, which I am creating my Bot for, but I guess you can imagine which one I'm talking about. I'll be replacing the name of the platform with "Scare".

The Beginnings
Starting this thread now feels a bit like cheating - Why do you ask?
Well, technically my journey started more than a year ago. I’ll elaborate.
I have been in the SMM Service industry on and off since I’ve been fourteen, but only really started earning money years later with a few B2C Shops. Money coming in always feels great, but Google Ads, Meta Ads, Customer Support, SEO and Shop design isn’t really for me. I am more of a technical person, so two years ago I started tinkering on my own Instagram and TikTok Bots (I even started a My Journey on those). In the end they didn’t end up working, but I had so much fun doing it, that a year ago I tried again with Scare (As I announced, this is the name I will be using for the platform).

The first Try
I had learned a lot in the year and got things working - kind of at least. Using Browser automation and some advanced fingerprint spoofing I was able to generate ~100 accounts and follow with them, but after that I was encountering more and more blocks. Frustrated I tried just using Requests and cutting out the Browser to more closely mimic real users and gaining more control over what to spoof and how to spoof it. After a few days I was able to create accounts, but they weren’t able to follow anyone. For a few days I kept trying, tinkering with settings, trying to make additional requests, changing proxies and more - Nothing worked! So I gave up. My main business was getting more time demanding again and I kind of forgot about the project for almost a year.

Then, about 2 month ago, I got sick. It was nothing serious, just a minor cold, but I was bound to my bed and went through my laptop and my Github, there I found the old project files and just gave it another go.
This time I was more persistent, I got the account creation going again after a day or two. Scare had made changes in the time, but they were minor and I managed to circumvent their systems successfully. But once again my accounts weren't able to follow other accounts. But this time I kept at it and after 10 days of trial and error, trying out different requests, changing useragents I managed to get this working.
I was jumping up and down through my apartment, being struck by joy. Unfortunately I can't share the exact steps I undertook to get it working, but let's just say it involved mimicing real user behaviour.


From Prototype to product
Having a proof of concept was amazing, but nothing more than a few Python files on my laptop, which allowed me to create trash accounts and follow with them. The accounts didn't have realistic usernames, weren’t using persistent sessions, didn't have profile pictures and I was controlling everything through a shell.
So I layed out a plan:

1. Add a Postgres Database to manage accounts, which users an account has followed, the email addresses which I used for registration
2. Add an UI to interact with my script, display accounts and statistics
3. Introduce Docker to my Setup. For anyone not knowing what Docker is -> It basically allows you to pack up your code into a neat little package, that can run anywhere where docker runs. It allowed me to test and develop my application locally and move it to production in a few clicks. It also allowed me to just spawn multiple instances of my application (workers) to handle account creation and following. Basically it made scaling easier.
4. Add Support for Profile Images, as the accounts looked a lot like bots^^
   I did this using S3 Buckets and uploading scraped Instagram Images to those. I then added logic to upload a profile image at random to each account.

All this was implemented until now. I have a clean Postgres Database, which stores a lot of info about the accounts.
The UI is clean and allows me to spawn Workers to create Accounts and also allows me to spawn follow jobs by entering a username and the desired amount of followers. As said earlier, I am not much a designer - so I let Codex do this for me. I think it looks clean!




How to monetize
We now arrived in the present, I have a working product - now I have to monetize it. My current goal is to offer a Follow Service to SMM panels primarily focusing selling Followers for targeted Locations (e.g. Turkey, USA, etc). I have the service attached to my own SMM Panel, which I have until now only used to run my B2C Shops, but I hope to sell the Scare Followers through it.
I observed that many of the followers being offered in Panels are trash, as they don’t have profile pictures, random strings as names and are being deleted after days.
Having followers which don’t look like bots, should be a selling point in my opinion.

The next steps
There are quite a few ideas that I still have for my project, as I am not yet 100% happy with the followers. After around 1-2 weeks some profiles incur problems with following. I am aiming to reduce this rate by making the follow process more realistic and mimicing the requests of Scare more closely.
Also currently the profiles don’t provide any other value than being “dumb” followers.
After understanding so much about the Scare API I want to implement a full viewer bot with sophisticated controls like Live Chatters, etc.
But for now I will try to get my Followers into some Panels by sending Cold Emails / Telegram messages.

Awesome plan, best of luck to you

More Accounts, more locations
During the last day I have created another 3000 accounts to increase my offering.
We are now sitting at:

• German Accounts total: 6578 (out of these 290 aren't able to follow anymore due to being shadowbanned)
• US Accounts total: 2613 (out of these 0 have been banned.)

I have already listed the Followers in my SMM-Panel and was able to sell Followers for around 5$. This of course isn't much, but I see it as a proof of concept.
To increase volume I created the US accounts to target a broader audience. Currently there is one other panel buying from me and two being interested.

To increase the quality of the service, I added parsing of the start count to ensure that Refills are possible based on the start count.
I also aim to implement a Cancel button that is wired into my services API to allow buyers to quickly cancel/partial orders.

Small improvements
There aren't any major news, but I just wanted to write a quick update on what changes I have implemented.
As I have already elaborated earlier, our Setup runs with connection to a SMM Panel (from Perfect Panel). Until now this was very basic as we just fetch the Pending orders every 30 seconds and provide updates to the Remaining Follows count, we also update orders when they are going into Progress in our System to the In Progress status and when they are finished to the Completed status.
In case a Username didn't have an affiliated Scare account, we of course cancelled orders, but there was no logic in place to cancel an order through the Panel.
Perfect Panel provides an API endpoint for that, so now every 30 seconds we don't only poll for new orders, but also Cancel requests. If a Cancel request is valid, meaning the order hasn't finished yet, it's accepted and the order is being cancelled. We then report that back to the panel and set the order Canceled/Partial.

We also had our first "bigger" order for just over 2.000 Followers. The system performed very well and delivered the 2.000 Followers in about 30 minutes. This might seem quite slow, but it's how I designed it to be. Orders shouldn't instant finish, but instead take a bit to appear realistic. If the orders become larger or we have a big order queue, the system automatically spawns a higher amount of Workers.

Current Revenue: 16$

I have read your topic completely.
You are a very cool person and it is interesting to watch the development of events. If this is the social network that I think about, then you will gain success in this direction of service.

Therefore, keep up the good work!

I would be happy to work with you if you have the desire for it and need some tests.

ALLSMMNET said:

I have read your topic completely.
You are a very cool person and it is interesting to watch the development of events. If this is the social network that I think about, then you will gain success in this direction of service.

Therefore, keep up the good work!

I would be happy to work with you if you have the desire for it and need some tests.

Click to expand...

Thank you for the encouraging words. I very much hope that the service becomes a success, as the project has been a lot of to me and I enjoy working on it.
Feel free to contact me, if you have questions on the project!

All the best for the journey.

More GEOs - More Money?

Well, I didn’t improve a lot in regards to our setup, but I added some more geos. Until now we only had German and USA Followers, that left a lot of locations out. I now added 1.000 Accounts for the Netherlands, Italy and France each. In the Panel I added these as own services + a EU-Service which unites all of the European Geos and chooses from the pool of accounts (because of that it’s 25% cheaper).



We haven’t seen any orders for that yet, but I’m sure it will improve the order volume.

As currently it sits at: 26.54$

Will follow along. Good luck brother. Sounds really interesting

This is pretty much how I got started too. Scraped some local business data for a friend who needed leads, realized I could do it faster than most people charge, and went down the rabbit hole from there.
What language you using? If its python and you're hitting rate limits, happy to share some stuff that helped me

Wow, what a great project and a great example of persistence, I'll be keeping an eye out for any updates you upload.

Some idea about the advanced fingerprint spoofing and also why accounts wont die +is your expense only on proxies or also other things

Project is awesome good luck

looks interesting keep updating