FREE SECRETGFX

TRAFFIC
FROM
YOUTUBE
SHORTS
AUTOMATED
 INTRO
In this guide, you will learn how to get free traffic from
YouTube shorts and how to automate the process with
python. There are countless different ways to make various
twists to this process as well as look for different niches and
monetization options. The intention of this guide is to
demonstrate how to achieve a fully automated & free traffic
source. It is  not  intended for you to copy the methods
shown as examples in this ebook. This is  not  a money-
making guide, but rather a well-researched traffic method
with monetization ideas. All the code shared in this guide
has been tested and working as of its release. There might
be changes to YouTube's algorithm in the future, so please
understand that you might need to adapt the methods
accordingly. YouTube does not like automated channels
and makes changes to the upload process often. In case the
upload script has stopped working, mention it in the forum
thread of this ebook. This guide took a lot of testing,
scripting, and general research to compile, please enjoy.
 Chapter 1
CREATING YOUR CHANNEL

hen it comes to short vertical videos YouTube does

W not care about the age of the account as long as you

can verify the phone number and email associated
with the account to unlock extra features such as
community posts and links in the description. However,
YouTube tries to copy TikTok and push the content to the
viewers from the country of origin. This means that if you
create a channel and upload a short video from Ireland,
over 90% of the viewers of the video will be from Ireland
and YouTube will try to push it locally as much as possible.
Luckily this is not TikTok and can be overcome fairly easily.
The recommended method throughout this guide is a
dedicated VPS. My personal, fully automated shorts
channel is running on a shared VPS with 1 CPU core and
2GB of ram. It takes a lot of time to render, but because it’s
automated, the speed of it does not matter to me. Virtual
servers with such specs can be had from as low as $5/mo. If
you are not willing to make any investments without
making money first, you can easily run everything on your
PC.
 Chapter 2
GETTING THE CONTENT

There are a million ways to get content for YouTube

shorts, but in this guide, I will show examples with
Cartoons and Reddit stories. Both of the example ideas can
get a lot of daily traffic, but these might not be ideal niches
for monetization. You can rather simply adapt the methods
demonstrated below to fit your own ideas and niches.

Example 1: Short Reddit horror stories

Example 2: Random funny cartoon clips

BEGINNING THE AUTOMATION

Step 1. You will need Python to run the automation. As
of the release of this guide, the most stable version for our
purposes will be 3.9.16 (not the latest one)

Step 2. Installing required libraries. To make things

easier we will install some Python libraries that will help us
with the automation process. To install a library just open
the terminal and type pip install library-name.
 Create a new folder for your project, type cd and
drag&drop the folder on top of the terminal window. Then
we can start installing libraries. It should look like this:

Here is the list of all the libraries we will need:

- MoviePy
- Requests
- Naked
- NLTK
- TTS
- BS4
Other requirements:
- eSpeak TTS module for Windows
- Some sort of text editor of your choice
- FFMPEG library for Windows
- ImageMagic editor software

Step 3. We will need some content to work with and the

first example is going to include Reddit. Not many people
know of this feature Reddit has. If you go to any subreddit
and type /random in the URL bar, you get a random thread
from that subreddit. This means that the system to get
content is already in place, we just need to ‘scrape it off’. I
personally like horror stories, so let’s try to create some
short horror clips using stories from /r/shortscarystories.

So we found the content we want, but we don’t want to

copy-paste it each time ourselves. Let’s write a short script
that does it for us. As of writing this guide, Reddit still has
their old mobile website active, which is super convenient
for scraping, as it is simple and loads quickly. We can access
it by typing i. In front of the reddit.com.

The main goal of this script:

1. Navigate to https://i.reddit.com/r/shortscarystories/random
2. Find the element of the body text
3. Extract the story text from the element
4. Save the story to file

We can easily find the needed element and their classes

using inspect tool in any browser. Here is how it looks:

As you can see the main body text of the post just sits
between some div tags with a class named ‘usertext-body’.
All we have to do is tell our bot to navigate to the webpage,
find the element with the same class, and get the text. For
this task we will be using Requests library to access the site
and BS4 to interpret the site’s HTML. Just like any popular
website, Reddit does not like bots crawling their page, but
luckily they are not that strict about it as long as you aren’t
causing too much trouble. Let’s use a random user-agent
just in case. If you want to scrape data in greater amounts,
you can do so by using their API but for this tutorial, we
will keep it simple. Keep in mind that the green lines with
hashtags are Python comments mean for the reader to
better understand the context. Here is how the code looks:

It should be saved with .py file extension and can be tested

by opening the terminal and typing py yourscript.py The
script should create a new .txt file in the same directory
with content similar to this:

Not the horror I was expecting, but ok

 Step 3. We have the content, let’s make a video! You
could do it manually, but that’s no fun, so let’s write
another python script that turns this into a video. This
video will consist of:
- Stock background footage
- Overlay story subtitles
- Text to speech VoiceOver
- Background music
For this demo, I will be using this background video, and
this background music. Now that we have everything we
need, let’s get started. The goal of this script is to crop the
stock background footage to a 9:16 aspect ratio, use TTS to
turn the Reddit story into audio, chop it into separate lines
of text that would fit on the screen, use TTS to turn these
lines into audio and overlay it with some background
ambiance. Sounds like a good recipe? The code for this task
is quite long, but don’t worry, it’s not that scary when you
read it. For convenience sake, I divided it into two parts:

jirencougar@gmail.com
At the beginning of this code, we define the necessary
libraries and download the NLTK module that helps to
divide our script into smaller sentences for subtitles to fit
the screen. Then we select the desired TTS voice. When
trying this code for the first time TTS library will download
the selected voice. This particular library has almost 100
voices to choose from and about 10% of them are very good
in my opinion. There are also quite a few languages besides
English. Here is the full list. Just change the model_name
number to choose the voice you want. The second part of
this code is fairly straight forwards, we just declare what
files we will be using to create this video and then cut the
background video to fit the vertical format. Let’s move on
to the second part of the code:

First, we begin by creating the subtitles, then set their

styling, overlay them over the main video, and fine-tune the
results by changing the clip speed and length. Some of the
stories in this subreddit will not fit into a 60 second short
video. There are multiple ways to solve it (find shorter
stories, make multiple parts, upload the full story as a
regular video, etc.) In this example, videos will be cut at
exactly 60-seconds to qualify as shorts. As you might have
noticed there is no code to generate subtitles. In line 88 we
call a function named subtitles, but I haven’t included it in
this screenshot, because it’s large and complicated. Think
of it like a car - you don’t need to know how the engine
works to drive it. I tried my best to comment every line to
make it as easy to follow as possible, but the only thing that
might need adjusting is the max word count per line and
subtitle speed. Here is how the function looks:

I marked the number that changes the subtitle speed (less =

faster), but to change the max word count per 1 subtitle line
you will need to change all the 4s to a desired number. The
current setting is max 4 words per subtitle line as with the
current font it’s difficult to fit more in a vertical video.
This is only meant to introduce the code and make it seem
more friendly for beginners, the complete and ready-to-be
copied scripts will be shared at the end of this guide.

End of example 1.
Now we have 2 script that can be easily joined to
generate endless short horror video stories. I do not think
this is a production ready copy-paste example, but I am
sure with some ideas of your own and slight tune of the
code you can auto generate amazing content that will get
thousands of views per upload. Curious to see an example
that was generated by the script above? Here it is!
You might be thinking, but how the content is endless if
you are just using one background video and same music
over and over again? That’s right, let’s continue on to
example 2 to learn about ‘content folders’ and how to
randomize elements to produce unique videos each time.

Example 2
This is the easiest way to get traffic from YouTube shorts.
I tried it multiple times and the views were overwhelming
each time. It might not be the ideal niche for monetization,
but traffic is traffic and you can figure out how to get some
money out of it regardless. In this example, I will show you
a script that cuts random clips from popular tv shows or
cartoons, makes it a bit more difficult to detect for the
YouTube algorithm (even though shorts allow some
copyrighted content), and finally add some text that
encourages viewers to check the comment section. There is
not much introduction necessary as the code is quite short
and simple. However, you will need a folder with .mp4 files
of your favorite tv show episodes to get started. Here is a
folder I have of some cartoon episodes:
 Now that you have a folder with some content ready,
let’s look at the code that turns it into shorts:

Don’t be intimidated by the size of this script, as 60% of it is just comments.

 Just like the previous script we start by importing the
needed libraries and choosing the files. Unlike the previous
example, we do not choose a single file, but rather a whole
folder and pick an episode randomly. You can use this
method for all the content from the previous example and
generate videos with random backgrounds, music, and
other elements picked by the script each time. It is also
possible to randomly scrape this content from other online
media sources. This eliminates the need to switch the
content manually each time. You can also write a simple
addition script that prevents duplicates as that might
happen sometimes. Getting back to the code, we take a
random .mp4 file and cut a short clip from it, excluding the
intro and end credits. Then we mix it up by mirroring the
clip and slightly speeding it up. After the clip is done we
overlay some text for our monetization needs and render
the final product. It’s barely noticeable that the video is
mirrored, cut, and sped up, most short viewers will enjoy it
regardless. Want to see a video generated by this code?
Here it is!
 Preparing for takeoff.
Just like everything you upload on YouTube, these videos
will need titles, descriptions, and tags to get those views.

We sure don’t have time to write it all ourselves every time,

so here is some code we can add to generate dynamic text
for our uploads. I would recommend adding as many
options for the title, description, and tags as possible, so
each generation would be somewhat unique and less
repetitive. Usually, you would want to max out the tags as
well for the best ranking. As you can see the majority of the
code is just arrays filled with options, which are randomly
selected and saved to file in the end. In line 5 we can see
that the title will always start with the word ‘Cartoon’
followed by a randomly chosen option from variable pos1,
then we have a # with a counter variable that returns the
number of our video, and finally the file name between the
brackets. Assuming the file is named “Simpsons S12”, a
possible outcome could be “Cartoon Deleted Scene #2
(Simpsons S12)”. The ‘\n’ command is added for a new line.
The same thing happens with the descriptions, except the
triple quotes enclosing the code mean that it is in multi-line
mode and no ‘\n’ commands are needed to change lines.
Adjust this as per your needs to build the desired dynamic
text generator. As you can see line 10 is a very basic attempt
to join 3 tags together, but usually, we want around 10
relevant tags to rank our video as well as possible. This can

be achieved with a simple loop like this:

You might notice that the last line of the description also
includes some tags, why is that? As previously mentioned
YouTube is trying to be TikTok and started to prioritize tags
that are in title or description instead of the ones buried
deeply in the metadata. This means that it is currently
beneficial to include tags not only in the dedicated spot,
but sprinkle some in the video titles or descriptions for that
ultimate ranking boost. Don’t forget that #shorts!

After a bit of practice and trying out both of these

examples, you should be able to make your script that
chooses the content and turns it into a short video suitable
for YouTube shorts. The quality of that video will mostly
depend on the materials chosen, but overall it’s currently
very easy to get traffic, even if your videos aren’t the best.
Some people do the Reddit stories with Minecraft parkour
backgrounds and TikTok TTS voiceovers. As we all know
those videos are terrible, but get millions of views anyway.
The sky is the limit, use your imagination and automatically
generate videos that can be uploaded multiple times a day
to get thousands of clicks.
 Chapter 3
AUTOMATING THE UPLOADS

So now we have a video made by a bot, but can a bot

upload it too? Sure can! Let’s create a script that uploads
our videos to complete this automation challenge. For this
method, we will be using NodeJS & Puppeteer instead of
Python & Selenium, because Selenium uses a web driver
that can easily be detected and is not welcome by most
websites, including YouTube. Puppeteer is not fool-proof
either, but I have been using this method for months and
it’s working flawlessly. Keep in mind that it is officially not
allowed to automate the upload process on YouTube, so I
am obligated to say that I am not responsible for any issues
caused. To begin, download NodeJS and install the required
libraries similarly to before. To begin this project, we will
need the Puppeteer library, which you can get just like this:
 Once that is installed, let’s grab delay-random library as
well. It’s not necessary, but very convenient and does
exactly what it says. When the node packages are in place
and you can see node_modules folder present, we can create
a new file named uploader.js and start working on the code.
Now please be warned that this is some messy, low-level
code, but due to how often YouTube changes the upload
process, it makes it easier to implement any changes:
 As you can see it just goes through the upload process as
a human would and uses keyboard navigation for the
elements it cannot reach by mouse. At the beginning of the
code, the upload_text.js is read and stored in an array line
by line. Assuming that line one is the title, line 2 is the
description, and the last line is the tags. The text is
accessed and typed by this command:
await page.keyboard.type(lines[0], {delay: 100});
Line 0 is the first line of the file, the delay simulates human typing.

Before we can try this code out, you will need the
upload_text.txt file we created using the previous script as
well as the cookies from your YouTube account, to access
it. Open your YouTube studio as you usually do and install
this extension then export all the cookies from the site. This
should leave you with a file called cookies.json. Now we can
take the upload bot for a spin. Open the terminal and type
node uploader.js to launch the script and begin the upload
process.
There are many ways to improve this code as it is as
basic as it gets, so that beginner readers would not have too
much trouble understanding what is going on with it. I
would not call this code very elegant either, it is just forcing
its way through the upload page, but I do not think it
matters as long as it achieves the desired action. Feel free to
improve and make this bot more sophisticated.
As you might have noticed I am not using a headless
browser, meaning that you will be able to see a browser
window opening and the script controlling it. Many people
would prefer using the headless mode to hide the browser
and let the action happen in the background, but
unfortunately, YouTube easily detects headless attempts
and closes them down. If you have a method to upload in
headless mode please share, but for general automation
purposes, this works just fine.
If you do not wish to add comments to all your uploaded
videos, you can remove everything from ‘#share-url’ to
‘upload completed’ and have a regular uploader script.
If you wish to change the video category or any other
upload options, add the necessary code after line 68. This
is a part of the code where the tags are entered and most
menus for category or language changes are accessible.
 Chapter 5
FINAL STEPS
Setting everything up
Wonder how to set everything up on a VPS to get passive
traffic going without any user input just as promised? Let’s
run through the final steps to fit all the pieces together and
reach that satisfying total automation. This will allow you to
keep the traffic coming even when you sleep, isn’t it
wonderful?

Step 1. Due to the upload script not working in a

headless mode we will need to trick Windows into thinking
that someone is logged in to prevent it from closing the
GUI. This can be easily done with an old utility called ‘RDP
Wrapper’ I am using an old Windows Server VPS, so I
cannot guarantee that it still works on the latest Windows 11
images, but most VPS providers allow you to choose an
older version for compatibility purposes anyway. Install the
software and try opening the configuration .exe file. This
piece of software
might trigger some
antivirus programs,
but it is a false
positive with no
harm. Set your
configurations to be
like this:
If everything was done
correctly you should
be able to open the
RDP window and
connect to your own
desktop ‘remotely’ just
like this:
By keeping this
connection active and working on the ‘remote’ computer
you trick Windows into thinking that there is always
someone logged in. Essentially, it just creates an endless
loop where one user is dependent on another user.

Step 2. Let’s pretend we are going with the cartoon clip

niche and we want to upload daily clips of random
cartoons. To make this happen we need to trigger our
script every day at a selected time. This can be easily set up
using the built-in Windows Task Scheduler. To be able to
launch a python script from this application, we will need
to create a helper script. Open a new text file and write this
batch command in a single line:
"C:\Users\Administrator\AppData\Local\Programs\Python\Python39\python.exe"
“C:\Users\Administrator\Desktop\folder-name\yourscript.py”

Adapt them as necessary to work on your machine.

Now save this file with a .bat extension, open the Task
Scheduler, and select ‘Create basic task’. Name it however
you want and select how often you want the videos to be
uploaded. When you reach a ‘start program’ page, browse
and select the .bat file we have created previously. Now
copy the path of the .bat file and paste it into the ‘Start in’
field without typing in the file itself. It should look like this:
 Once that is set, save the task and adjust the timing as
needed, there are plenty more detailed tutorial on YouTube
as this guide is not about the Windows Task Scheduler.

Step 3. Now we have a python script that will get

triggered once (or more times) a day automatically and
generate video on it’s own. But our upload script is on a
separate file in a different language, how do we join them?
To do this, we tell the python script to launch the uploader
script after the video is ready, by using this simple
command:
execute_js(‘uploader.js')
 It’s really that simple! However, for that command to be
accessible you do need to include the Naked module first.
from Naked.toolshed.shell import execute_js, muterun_js

Now every time the video generator script is launched, it

will prepare a new video and automatically call the
uploader to upload it.

Conclusion.
Make sure you have read through the code and are
familiarized with the commands. This might be a bit tricky
to orchestrate for the first time, so here is a flow chart that
should help you understand better:

Task Manager set to launch start.bat at 1 am ->

Time is 1:00 -> start.bat is active ->
The batch command triggers cartoon_short.py ->
Video is generated and saved as clip.mp4 ->
Video title & description generated and saved as desc.txt ->
Python is calling NodeJS uploader.js script ->
uploader.js looks for cookies.json & desc.txt ->
uploader.js launches Chrome and starts uploading ->
uploader.js finishes the upload and adds a comment ->
uploader.js closes -> cartoon_short.js closes -> task finished.

I am sure that there are many different ways to do this, but

the basics are here for you to grab and apply. This should
not take longer than an hour to set up in total and the
traffic you will get passively should be truly impressive.
 Chapter 4
SHARING IDEAS

Let’s begin this chapter by talking about the frog 🐸 in

the room - Coqui TTS. It’s the python TTS library we have
used in the first example. This library has plenty of options
and settings to play around with to get that perfectly
realistic human voice out of a single file on your computer.
I agree that the voice in the example is not perfect, but if
you take a moment to read through the library's
documentation and learn how to adjust and fine-tune the
details as well as import custom voice packages from other
creators and alter word pronunciation or use multiple
speakers at once. I would also like to remind everyone that
there are more languages than just English and some
content might be even easier to rank and get views in other
languages. If you have an Nvidia GPU you can also use your
CUDA cores to speed up the TTS generation and render
videos in seconds.
Now you probably noticed that the monetization
examples in this guide are not the best. However, this guide
is about teaching how to get large amounts of free traffic. It
is up to the individual reader how to monetize or utilize
that traffic. That said, I will share one beaten-to-death idea,
that still somehow works. Game hacks. With a tiny bit of
brain power, you can easily adapt the scripts from this
guide to generate endless short game hack videos with
download (CPA) links in the comments. There are plenty of
subreddits sharing modded APK files and there is already a
way for you to get those posts. All left to do is add some
extra elements to make the videos more appealing to a
younger audience and you are making money from the
automated YouTube Shorts channel! I tried it myself and it
works, however, I would only recommend this niche if you
are desperate, because the competition is massive.
I bet there are at least a million niches that could be
used in conjunction with this automated shorts method.
Take your time to test the scripts out, get familiar with each
line of code and adapt it to any niche to make that money.

How much traffic can you get?

Here is a screenshot from a channel I created 1 month
ago. All the content uploaded is just movie clips using the
method from example #2. I only logged in to this account a
few times to check the stats. All the traffic was made
completely passively with no input from me. I did try to
include a link in one video but have not made any effort to
promote it. You can see the clicks in the first graph. This
channel is no longer active as I am working on different
niches right now, but it is still getting decent daily views.
 I truly think that this shows how much it can be scaled.
Imagine creating 10 channels that are completely
automated and easily getting 1000 clicks to your CPA offer
every day. This is possible if you take action and invest the
time needed to set everything up in a sensible way. Don’t
skim out on the content either. It does not mean that
automated videos have to suck. Some of them can truly be
entertaining and valuable to the short community.
There is also always a possibility to get monetized by
YouTube directly. All you have to do is fine-polish that
robotic voice to sound more like a human, spend more time
cleaning up the raw content to fit the video format better,
and reach a level of production that might be deemed
acceptable by YouTube gods.
Let’s not forget TikTok and the massive user-base there.
I do now have a reliable method to share on how to upload
content automatically to TikTok, but I am sure that some of
the scripts from this guide can be adapted to automate
traffic from any platform.
If you come this far into this guide, thank you for
reading and I hope it was a valuable piece of information.
You will find all the code from previous examples below.
Feel free to copy and test it for yourself. If you have any
issues check out the troubleshooting page or contact me
directly on BHW. Now go make that traffic start flowing!
 CODE DUMP

Individual pieces of code:

Scraping Reddit for content -> Pastebin
Generating videos from stories -> Pastebin
Cutting clips from tv episodes-> Pastebin

Complete scripts:
Reddit to video -> Pastebin
Cartoon clips -> Pastebin

YouTube uploader:
Video upload -> Pastebin
 TROUBLESHOOTING
Q: My code doesn’t work, what do I do?
A: Paste it in chatGPT and ask what is wrong with it.

Q: My code still doesn’t work, what do I do?

A: Contact me on BHW and I will try to help you

Q: It generates the video, but doesn’t upload

A: Make sure all the required files are in the same folder

Q: I installed ImageMagick, but it still shows it’s missing

A: Select all the checkboxes when installing

Q: I installed FFMPEG, but it still shows it’s missing

A: Try a different, older version, like this one

Q: How do I install FFMPEG, it’s just a bunch of files?

A: Like this

Q: I installed *thing* but it’s not working

A: Have you added it to PATH? Have you tried restarting?

Q: Can’t upload, says browser not supported

A: The cookies are corrupted or missing

Q: The code provided doesn’t work (moviepy error)

A: Wrong version of Python or missing dependencies