When I streamed in 2020, I ran the same show on both Twitch and YouTube. On Twitch, with zero promotion, talking about propaganda, psychology, art, journalism, and other outré topics, I’d see anywhere from a few hundred to a few thousand viewers, and the numbers climbed over time. That’s modest, but credible for the niche. YouTube was another matter.

On YouTube, I would watch the view number more than halve in front of my eyes. I’d end up with more likes than views. I’d get more DMs from people discussing the stream than the total views YouTube claimed I had. I left streaming. When I came back for KlueIQ, the same thing happened again, only worse.

Fast‑forward to now. My latest True Crime, Wrong Story O.J. Simpson episode aired three days ago. Today, YouTube shaved it from 25 views down to 24. This is not new. I have never trusted YouTube’s numbers. They have always reeked of a scam.

So I conducted an experiment: I asked people from different countries to watch that particular episode: no ad blockers. They all have different IPs, devices, and knew the parameters and proof of participation. This was after the original aired. Then I went to look at the episodes with different browsers to see what would happen.

Assuming no one else watched, the numbers should have been 57 if none of my control group viewings were counted, and 105 if they did.

The baseline was 26 before the experiment.

So, if YouTube numbers were accurate, it should have been 57.

If they were promiscuous: 105, but given the subject matter, it should be much more than that from organic traffic.

Not 24.

So what’s going on?

There are two theories:

1. I am on some banned list, where the point is to personally humiliate me with low numbers and discourage me from putting out content and viewers from watching.
2. There is a hidden financial incentive to play these games.

The second explanation makes more sense because it is not just me we ridiculously low numbers, but many other very good channels.

If we look at Alphabet’s business model, we can glean a lot of clues, and it boils down to the difference between upfront payment structures and consignment structures.

So let me backtrack to my early adulthood where my mother made beaded jewellery: first as a hobby, then as a designer selling her work in various stores from Windsor to Toronto to Niagara Falls. It was a large territory (Hamilton, Ancaster, Waterdown, Burlington, Oakville, and Niagara-on-the-Lake being the cities with the most stores and presence).

The stores at the time either bought her jewellery outright (upfront payment) or took pieces and then paid once they sold (consignment).

I was the one who picked the stores and dealt with them. The results were very different, yet predictable:

1. Those who paid upfront wore the pieces, displayed them prominently, asked for custom orders, and asked for more because they sold out and regularly. I had good to exceptional relationships with each one and there were never any “stories” about lost merchandise and nothing was ever returned to us as defective.
2. Those who worked on consignment never placed the pieces in prominent places, and often “forgot” and kept them in a drawer. There would be returned pieces that were damaged, and every place had at least one convoluted and dramatic “story” about how an expensive piece was stolen, and we were never compensated. This second group were mostly dropped from our list.

To put this in perspective, my mother never put tags on her pieces. No “Made in Canada” selling point, no designer label, and she refused to be tethered by doing “lines.” She still sold incredibly well based purely on style and quality; this wasn’t her job, but a reliable, profitable, gratifying side gig.

When a store owner had skin in the game, they bought the pieces up front, the jewelry flew off the shelves. They wore it, displayed it, reordered it. When they took it on consignment and got the product for free, suddenly pieces “gathered dust in a drawer.”

Or so they claimed.

Because if you’re not upfront about what you’ve actually sold, you can pretend you didn’t sell anything and hold off paying the maker. If an expensive piece happens to go “missing,” you can quietly sell it under the table to a special customer, pocket 100% of the profit, and the vendor is out of luck. The seller controls the shelf, the story, and the till.

YouTube made its fortune as a consignment publisher.

Creators hand over finished work for free. YouTube decides where to shelve it, how to package it, and what to report back. If they choose to classify a chunk of your audience as “invalid traffic” or simply undercount it, your view count, and therefore your pay, shrinks, but you have no independent way to audit what actually happened.

And yet, YouTube is fully capable of being an upfront publisher when it wants to be.

They bought the exclusive global rights to broadcast the Oscars starting in 2029. That’s not cheap, but Alphabet’s deep pockets can handle it with ease. The Academy gets a real deal and real guarantees. Everyone else? We have no idea.

Even large, popular channels have no way of knowing whether the numbers they see are the whole truth. YouTube can show you graphs and geos and devices, but if they decide to suppress or filter, they can show advertisers one reality and creators another, or quietly move the goalposts in the name of “invalid traffic” and “quality control” while keeping the upside.

If you are a creator, I strongly suggest running your own experiment.

Ask a known group of real people, different devices, different IPs, no ad blockers, to watch a specific video and send you proof. Count the humans. Then compare that to what YouTube claims. There is a glut of YouTube advice channels telling you what you are doing wrong; almost none of them question the metrics themselves.

Meanwhile, legacy publishers are finally turning their fire on Google’s algorithms and the supposed loss of traffic, but they still rarely ask the blunt question: is Alphabet suppressing or selectively inflating numbers to game the system in its economic favour?

I don’t think that’s a conspiracy. I think that’s a business model.

Noticing that numbers don’t add up is not a conspiracy theory. It’s basic numeracy. The theory comes later, if at all, and I’m starting with the math.

If your lived experience and YouTube’s numbers don’t match, don’t gaslight yourself. Run an experiment. Count your humans. Compare that to what the platform says. Treat the view counter as a claim that needs to be tested, not an oracle to be obeyed.

If you try a controlled test like mine, I want to hear about it (DM me via LinkedIn or Facebook), especially if you can document:

• How many real people watched (with proof).
• What YouTube reported 24–72 hours later.
• Any sudden drops, “corrections,” or views disappearing in real time.

I’m not asking you to accept my story. I’m asking you to compare your own math to YouTube’s and then ask the same question I am:

If the numbers are always right, why are they so afraid of being checked?