My YouTube Journey: From Zero to 3 Million Views As an Introverted Guy From India

It all started way back in the year 2025, in the 4th quarter to be precise. I was an extremely introverted guy back then. But I started taking YouTube quite seriously, because I knew one thing: if I kept rotting inside my room as an introvert, then conditions would turn from bad to worse. I had a childhood fear of speaking in front of the public. Today, when I look at my channel’s analytics dashboard, I feel that the success would not have been possible if I had not stepped out of my comfort zone.

I can confidently say that I have overcome:

1. Fear of public speaking
2. Fear of talking with strangers
3. Fear of taking risks

Grind, Sweat & Tears

My Story in a Nutshell: I casually started making videos on YouTube regarding Higher Education Scholarships, Hackathons, Competitions, Tournaments, Contests, Tips and Tricks, Public Awareness, Exciting Opportunities, Open-source Contribution Guide, Critical Thinking and Brainstorming, Videos On-demand and much more.

1. For many weeks, I’ve been through burnout from inside-out because of intense Video Editing I would do myself. From topic researching-to-script writing (well, I don't write scripts, and I speak whatever comes into presence of my mind at a moment)-to-screen recording( sometimes, video capturing)-to-editing on a video editing software of old times.
2. From the last week of August 2025, videos started finding its audience. Basically, school/college/university students, young minds, curious people, and other people who think of innovative ideas, is my audience base. The Age bracket is 18-25 years, as per the stats. I got so many hearts not just from YouTube but also from people on Reddit admiring my work, passion and zeal. Yes! I have built a Reddit community too. As of now, it is 2,300+ members strong. I’ve seen my audience demographics. I’ve seen an impact. Most importantly, I have been making videos to keep myself busy, just not to left with loneliness & mood swings.

A solid piece of advice for aspiring content creators:

1. Subscriber base doesn’t matter much unless your content is highly engaging.
2. Choose a specific niche. Otherwise, be prepared to wait for years if you plan to deliver multi-niche content, just like me.
3. Try solving real-world problems. Make guidance videos, tutorials, troubleshooting content, tips, and tricks. This builds confidence and a loyal audience organically.
4. Spend some time on research, video recording, graphics, thumbnails, editing, and spreading the word about what your channel is all about. At least keep 12 hours of the day aside for this work if you’re doing everything single-handedly.

Now, I envisioned helping others in the best way possible so that I could contribute back to society (which is the world, of course).

To start a YouTube channel, all it requires is:

1. A Google Mail account (better if you create a separate account instead of using your personal one, as it will help you in the long run)
2. A problem-solving mindset
3. Fear and desire (both)
4. Perseverance

Unless you’re trying to become a second version of MrBeast, you won’t need a huge sum of money. In fact, I have single-handedly grown my YouTube channel.

At the very beginning of my YouTube journey, I used to follow a standard advice that tells you to "pace your video" so the viewer does not get overwhelmed.

This is dead wrong.

If a viewer understands everything effortlessly, they enter a "passive state." Passive viewers drift. Passive viewers click off.

I have been tracking the "post-interaction velocity" of some of my viral videos.

Look at the picture. See that spike?

That is a "seek_back" event.

When a user manually rewinds your video to see something they missed, the algorithm flags that timestamp as a "viral moment." It assumes the content was so valuable or dense that it required a second look.

The engineering exploit: "the visual density flash."

You need to intentionally overwhelm the viewer’s processing speed once per minute.

The artifact: create a dense resource (a tier list, a spreadsheet, a hidden detail, a map).

The flash: put it on screen for 1.5 seconds.

The result: the human eye takes ~0.8 seconds to recognize the image but needs 3+ seconds to read it.

The viewer thinks: "Wait, that looked important."

They rewind. They pause.

The math:

To the algorithm, this is "active engagement."

A video with 50% retention and zero rewinds is "good."

A video with 50% retention and 500 rewinds is "god tier."

We call this the endowment effect. When a user works (rewinds) to get the information, they value it more.

Stop spoon-feeding your audience. Make them work for it.

Over 90% of so-called Gurus (it is a Sanskrit term for a "mentor, guide or expert" in a certain knowledge or field) keep saying, "Just keep uploading! Consistency is key!"

1. This is mathematically dangerous advice in 2026.
2. I am at a stage where I can confidently say that the YouTube algorithm does not reward output. It rewards outcome.
3. Specifically, it tracks a metric called "channel authority" (trust score).

The engineering reality:

1. Every time you upload a video that gets low retention (AVD) and low click-through rate (CTR), you are feeding the neural network a negative data point.
2. If you upload three times a week, and all three videos perform poorly, you are aggressively training the algorithm to ignore you.
3. You are building a "confidence interval" that says: this channel consistently produces low-satisfaction content.
4. Once that confidence interval is established, the algorithm stops testing your new uploads. You are "shadowbanned" not by a human, but by your own bad data.

The frequency trap:

1. Consistency only works if the quality is above the niche average.
2. Good video + consistency = exponential growth.
3. Bad video + consistency = exponential decay.

The fix: the quality pause.

1. If your last three videos flatlined, stop uploading.
2. Do not "push through." The algorithm is not testing your work ethic; it is testing your efficiency.
3. Take two weeks. Re-engineer the packaging (thumbnails) and the hook (retention).
4. Upload one video that hits 60% retention.
5. That single positive data point outweighs ten mediocre uploads because it breaks the "low satisfaction" pattern.

The hard truth:

Nobody cares how hard you work. The machine only cares about the signal. If your signal is noise, turning up the volume (uploading more) just makes it annoying.

Stop grinding. Start optimizing.

Stop trying to be "unique."

1. You know, there are videos where I tried to be "original" and "creative" just flatlined completely.
2. Here is why your videos are stuck at 50 views based on the actual engineering of the system.
3. The algorithm is literally just a "nearest neighbor" machine (k-NN). Google does not actually watch your video. It converts it into a string of numbers called a vector and places it on this giant infinite map. The logic is basically k-NN (k-nearest neighbors).
4. When someone finishes a MrBeast video, the AI just scans the map and asks, "Okay, which video is the mathematical neighbor to this one?"
5. If you are the "unique" creator making a video about a topic nobody is talking about, your vector gets put in "the void" (empty space on the map). The AI looks for neighbors, finds literally nothing, and gives you zero impressions.
6. But if you are the "drafting" creator, you make a video that is mathematically similar to one that is already viral. You park your vector right next to the "gravity well," and the AI sees you are a neighbor and drafts you into the stream.
7. The lesson is the algorithm is not a discovery engine; it is a prediction engine based on proximity, not quality.

To use "vector drafting," you need to stop trying to invent new topics and start "drafting" existing vectors.

How:

1. Find a video in the niche that is popping off right now. Target those specific coordinates (topic + keywords).
2. But the pivot is key: do not copy them. Position your video as the sequel or the answer to their video.
3. If they made "Why Bitcoin is Crashing," you do not make "Why Bitcoin is Crashing." You make "Why Bitcoin Will Bounce Back (The Math)." You stay in the same vector cluster but you offer new information.

The results? The channels using this strategy will grow faster than the "creative" channels.

Originality is a luxury you earn after you have an audience. In the beginning, you just have to respect the math. You must be a neighbor.

My channel crossed over 3.3 million views in the last 28 days with just 6k+ subscribers. In fact, all other metrics are sky-high. I made use of Ask Studio, an in-built feature on the YouTube dashboard for content creators. Hear me out, understanding the dynamics of the YouTube algorithm is not rocket science. Believe me, no one has ever shared such raw advice in the history of mankind.

Yesterday, I audited some of my videos that have great retention until the last 20 seconds.

Then, the graph falls off a cliff.

Why? Because I said the "exit words."

"Anyway guys, that’s it for today..."

"Thanks for watching..."

"Don’t forget to like and subscribe..."

The engineering reality:

These phrases are audio triggers.

When a viewer hears them, their brain subconsciously decides: "The value transaction is complete. I can leave now."

They close the app.

To the algorithm, this is a session termination event.

The system flags your video as the "stop sign" that ended the user’s session.

If you consistently end sessions, the algorithm stops recommending you because you lose them money (ad revenue stops when the user leaves).

The goal:

You don’t want to be a "stop sign." You want to be a "bridge."

The algorithm rewards videos that generate downstream views (session extension).

If a user watches your video, then immediately watches another one, you get credit for that second view too.

The fix: the bridge technique.

Stop acting like a TV host signing off. Act like a dealer giving the next hit.

1. Delete the outro: no black screen, no scrolling credits, no "goodbye."
2. Open a new loop: in the last 10 seconds, do not summarize the current video. Introduce a new problem that is solved in your next video.
3. The hard cut: point to the end screen card and cut the video to black mid-breath.

Example:

Bad (polite): "So that’s how you fix audio. Thanks for watching, see you next week!" (user leaves).

Good (engineered): "Now you have good audio, but your lighting still looks terrible. The secret to cinematic lighting is actually this $20 tool right here..." [points to card → video ends].

The user clicks because they have a new problem. The session continues.

Stop saying goodbye. Nobody cares. Just give them the next video.

Most editors think "retention" is just about being loud or fast or whatever. It’s not. It’s actually biology.

Hear me out, retention drops are rarely caused by the "content" itself. It’s usually just biological fatigue.

Here are the three bio-hacks that literally define the difference between a 30% AVD and a 70% AVD.

1. The "0-degree" rule (saccadic cost): the human eye moves in rapid jumps called "saccades." Every jump costs energy (ATP). Your brain is lazy, so if you force it to spend energy just to find the subject, it gets annoyed.
2. The mistake (high cost): let’s say in scene A the subject is center. Assume in scene B the subject is top right. The viewer’s eye has to physically hunt for the new spot. This creates "visual friction."
3. The fix (0-degree editing): you have to align the focal point of the new clip to match the old one.
4. The result: the eye stays still while the world moves around it. This lowers the "cognitive load" to zero and basically puts the viewer in a trance.
5. The "dopamine pulse," which is characterized by a biological reaction during scene changes: pupil dilation.
6. When the visual environment changes, your brain releases a micro-dose of dopamine to process the new information. This creates a "time dilation" effect where the viewer loses track of time.

But the brain builds tolerance fast (habituation). You need a specific frequency to keep the pupils dilated:

1. Long form (the 7-second rule): every seven seconds you have to change the "state" (camera angle, b-roll, text overlay). If it stays static for more than 10 seconds, the pupils constrict and the brain checks the sidebar recommendations.
2. Short form (the 1.5-second rule): for TikTok/Shorts the tolerance is much lower. You need a visual pulse every 1.5 seconds to stop the "swipe reflex."

The "physical disarmament" (the snack): this is a psychological trigger for long form.

In the first 60 seconds explicitly tell the viewer: "Grab a drink or a snack, sit back, and listen."

Sounds polite, right? Actually, it’s a trap.

The logic: if a viewer is holding a sandwich, both hands are occupied. They are physically unable to touch the mouse to skip.

The result: you force them into "lean back" mode. By the time they finish the snack, you have bridged them deep into the content.

In my videos, I often tell students (most of my audience is school and college-going students) to take a copy and pen because I make videos in the education niche—public awareness about education funding opportunities, to be precise.

Conclusion:

The algorithm is just code designed to mimic the human brain. If you edit for the eye (0-degree rule) and the chemicals (dopamine pulse), the algorithm will follow. Trust me.

More often than not, I see beginners every day screaming, "YouTube hates me" or "I’m shadowbanned" because impressions dropped to zero after 50 views.

Let me clear this up: the algorithm does not have emotions. It does not "hate" you.

It just does not trust you yet.

When you upload a video, the neural network gives you a "seed audience" (usually 100–500 impressions) to test its prediction.

If you get a 5% CTR? Great. The system trusts the data and pushes you to the next batch (1,000 people).

But if you get a 2% CTR? The system does not think "this video sucks." It thinks "my prediction was wrong."

To save server costs, it stops showing the video immediately to prevent further "bad predictions."

To you, this looks like a shadowban. The switch just flips off.

To the engineer, this is just efficiency.

You are not being suppressed. You just failed the initial verification test.

New-age content creators highly misunderstand this.

After making nearly 500 videos (both short & long form), I have realized that machine learning systems often exhibit an implicit bias towards the past.

This is the smoking gun.

It means the natural state of the algorithm is to ignore your new video. The AI is designed to be "safe," so it prefers recommending videos that already have 10 million views (the past). To the math, your new upload is a risk.

So how do new channels ever grow?

Google engineers had to manually hard-code a "freshness" variable to force the AI to give you a chance. This is what we call the "freshness boost." For a short window of time, usually 24 to 48 hours, the algorithm artificially inflates your probability score to fight its own bias against you.

This is why your retention graph flatlines.

You are not fighting a shadowban. You are fighting a mathematical decay curve. If you do not generate massive velocity (high CTR and engagement density) during that tiny "freshness" window, the model reverts to its default state which is ignoring you.

Stop uploading and praying.

If you do not spike the graph at hour zero, you are wasting the only help the code is ever going to give you.

When a video gets a burst of views and then suddenly flatlines, it feels like someone flipped a switch, right?

Well, it used to happen with my videos quite often.

The algorithm has a natural bias towards "old" videos because they have more data. It is safer to recommend a MrBeast video from 2022 than your new video from today.

For the first 24–48 hours, the algorithm artificially boosts your probability score just because the video is new. It is giving you a "free pass" to get noticed.

The problem:

1. Most small channels think the flatline is a "shadowban."
2. It isn’t. It is just the freshness boost expiring.
3. If you did not get enough clicks and watch time during that free boost window, the algorithm removes the training wheels. Your probability score drops back to zero, and the video stops moving.

How to fix it:

You cannot rely on "luck" during the first 24 hours. You have to engineer velocity.

1. Priming: don’t just upload cold. Use the community tab to wake up your subscribers before you drop the video.
2. The click: if your click-through rate (CTR) is low during the boost, the boost ends faster.

Don’t be discouraged when the views stop. If you win the sprint, you get the marathon.

Hope this explains the "mystery." It’s just math.

I see a lot of channels every day pasting a block of 50 keywords at the bottom of their description, thinking it helps SEO.

I honestly laughed so hard seeing this, mainly because most creators are still stuck in 2015. They think "SEO" just means stuffing keywords in the description box like "funny moments, Minecraft, gaming, viral, 2026" lol.

It doesn’t. In fact, under the modern architecture, this is likely flagging your video as "low quality."

The engineering reality:

1. In 2012, YouTube matched "keywords." If you typed "Minecraft," it looked for "Minecraft."
2. In 2026, YouTube uses natural language processing (NLP), specifically a model called BERT (Bidirectional Encoder Representations from Transformers).

The system does not read words in isolation. It reads context. It tries to understand the "relationship" between words to determine the "semantic value" of the video.

Why keyword stuffing fails:

When you list: "funny, viral, Minecraft, gaming, let’s play"... that is a "low entropy" data string. It has no grammar. It has no context.

In fact, I took this experiment to the next level by opening a competitor’s video on Chrome browser on my Lenovo laptop. Then, pressing Ctrl+U. Then, Ctrl+F to search for "tags" to see their metadata

Believe me, this looks like noise. It lowers your "trust score" because humans do not speak in tag-lists. Only bots do.

The solution (the description pyramid):

1. You need to stop writing for humans and start writing for the NLP model.
2. Delete your tags: YouTube engineers have publicly stated they have minimal weight.
3. Use natural language questioning: the AI is trying to answer a user’s intent.
4. Bad: "how to bake cake, cake recipe, baking.
5. Good (NLP optimized): "In this video, we answer the question: how long should you bake a cake at 400 degrees? We also discuss why your cake collapses in the oven."

This gives the AI the question and the context. It matches the user’s search query perfectly.

I call this semantic indexing.

Some of my videos even have 60% retention but zero comments.

I questioned myself: "My video was perfect, why didn’t it blow up?"

Answer: because it was perfect.

..because I realized that when you are trying to be Wikipedia, nobody comments on Wikipedia. They consume the information and leave.

That is a passive session. The algorithm hates passive sessions.

To go viral, you need active engagement density. You need people typing.

Here is the psychology hack:

Cunningham’s Law: "The best way to get the right answer on the internet is not to ask a question; it’s to post the wrong answer."

Humans have a biological need to correct others. It signals status: "I am smarter than the creator."

The engineering exploit (the tripwire):

Stop making your videos 100% airtight. You need to plant a tripwire—a specific, harmless error designed to trigger the "correction reflex."

1. The typo: misspell a word in a text overlay for two seconds.
2. The mispronunciation: say "GIF" wrong. Call a "magazine" a "clip."
3. The ranking: put a popular item in "C-tier" on a tier list without explaining why.

The math:

When 500 people rush to the comments to type "UM ACTUALLY IT’S PRONOUNCED…," the algorithm does not see "hate."

It sees velocity.

It sees time-on-page.

The algorithm is sentiment-agnostic. It does not care if they are correcting you or complimenting you. It only counts the engagement points.

The rules:

1. Never error on the value: the core advice must be good.
2. Only error on the details: the tripwire must be trivial.

If you are perfect, you are boring.

Give the audience a reason to feel smarter than you. Let them correct you. You get the views; they get the dopamine. It’s a fair trade.

Stop being a professor. Start being a provocateur.

My heart palpitated every single day waiting for a letter from Google.

The final verification letter which comes at your door-step when your YT channel crosses a benchmark of 4k watch hours and 1k subs within last 365 days.

It consists of a 6-digit PIN.

Well, I make videos out of passion. My side-hustle paid off.

The postman smiled (as if he knows what's inside) and gave my dream letter.

Your YouTube earnings are paid through your AdSense for YouTube account, which requires you to reach a Payment Threshold before a payment can be issued. If your balance does not meet this threshold, your earnings roll over to the next month until the minimum is reached.

The Payment Threshold varies based on the reporting currency of your AdSense for YouTube account:

Currency Payment Threshold:-

1. U.S. Dollar (USD) - $100
2. Euro (EUR) - €70
3. Great British Pound (GBP) - £60
4. Indian Rupee (INR) - ₹8,000
5. Canadian Dollar (CAD) - C$100

Payment Process

The payment process follows a monthly cycle, and to receive a payment in a given month, you must meet the payment threshold by the 20th of that month.

1. Between the 7th and 12th of the month: Your estimated earnings from the previous month are finalized and posted to your AdSense for YouTube account.
2. On the 20th of the month: Your total balance must meet the Payment Threshold. Any changes to your payment information, including removing payment holds, must be completed by this date.
3. Between the 21st and 26th of the month: Payment is issued. The exact time you receive your payment depends on your chosen payment method:
4. Electronic Funds Transfer (EFT): Allow up to 7 business days for the transfer to complete.
5. Wire transfer: Allow up to 15 business days for the transfer to complete.

Other Revenue Thresholds

There are other one-time revenue thresholds you must meet to complete the necessary steps for your first payment:

1. Verification Threshold (usually $10 equivalent): This is a one-time minimum earnings requirement. Once your finalized YouTube revenue reaches this level, YouTube/AdSense will need you to verify your identity and your address. This is typically done by uploading an image of a government-issued ID and then receiving a PIN mailer at your physical address. Completing this step confirms your information and is mandatory to continue your channel's monetization and get paid.
2. Payment Method Selection Threshold (usually $10 equivalent): This threshold often aligns with the verification step. Once you reach it, you will be able to select your preferred payment method (like Electronic Funds Transfer or wire transfer) in your AdSense for YouTube account settings. This sets up the bank details for when you finally meet the main Payment Threshold and a payment is issued.

Think of them as early, crucial steps to set up your account so that you're ready to receive your full earnings once you hit the main Payment Threshold (e.g., $100 USD).

Here are some types of questions, categorized by research focus, that you can ask to get a deep analysis of your channel's metrics:

1) Audience Deep Dive

Understanding who your audience is and what they want is the foundation of content strategy.

1. Demographics and Viewer Loyalty:
2. "What is the breakdown of my audience by age and gender?" (We will look at the last 28 days by default).
3. "What percentage of my viewers are subscribed versus not subscribed, and how does their watch time compare?"
4. "How many of my viewers are brand new to the channel, and how many are returning for the last 90 days?"
5. Audience Activity and Interests:
6. "When are my viewers typically online?" (This helps with optimal posting times).
7. "What search terms is my audience using on YouTube?" (This is great for finding new topics that your audience is actively looking for).

2) Content Performance & Optimization

Analyze which videos are resonating and why, so you can double down on successful formats.

1. Top and Bottom Performers:
2. "What are my top 5 videos by Views from the last 90 days?"
3. "Which videos published in the last 28 days have the highest Click-Through Rate (CTR)?"
4. "Which video content type (Shorts, long-form, or live streams) had the most Impressions in the last 30 days?"
5. Video-Specific Deep Analysis:
6. "How does the average view duration for my video [video ID] compare to my typical performance for that content type?" (We compare it against your last 10 similar videos).
7. "Show me the audience retention curve for my video [video ID] and analyze the biggest drop-offs and spikes in the first two minutes." (We'll use the video's captions to figure out what was happening on screen at that moment).
8. "What are the main topics or themes viewers are commenting about on my video [video ID]?"

3) Channel Health and Discovery

Look at how your channel is growing and how viewers are finding your content.

1. Growth and Publishing Trends:
2. "Can you show me the trend for Views and Watch Time over the last 90 days?"
3. "How has my publishing consistency and growth in new viewers changed over the past year?"
4. Discovery Analysis:
5. "Where are my long-form video views coming from? What are the top traffic sources like Suggested Videos or Browse Features?"
6. "What are the top external websites that are linking to my videos?"

4) Monetization Metrics

If your channel is monetized, you can research revenue patterns.

1. "What was my total revenue and RPM (Revenue Per Mille) for the last 28 days?"
2. "How does my revenue compare period over period, for example, last 28 days versus the previous 28 days?"
3. "What percentage of my revenue comes from Shorts Feed Ads versus Watch Page Ads?"

My Channel's Revenue Breakdown (as of February 14, 2026)

1. Long-Form Videos (Videos): 86.63%
2. Shorts: 9.48%
3. Live Streams: 2.64%

By asking these detailed, metrics-driven questions, you can move beyond simple performance tracking and truly use your data to inform creative decisions and drive channel growth.

Wish you success!