Channel analytics: the six numbers that decide whether there will be money

Updated 06.10.2026 · Kostas Obuchow

Subscribers and views are not among them. The six numbers that actually move revenue, where each one lives in Studio, what a bad value looks like, and the difference between CPM and RPM that costs people the most confusion.

Short answer

Subscribers and total views are not on this list. They are the two numbers everybody reports and the two that predict revenue worst — a 400,000-subscriber channel can earn less than a 20,000-subscriber one, and routinely does.

Six numbers decide it:

Numbers 3 and 5 are where almost all of the surprises live, and they are the two people check last.

1. Average view duration

This is the number the recommendation system reads first, and it is also the one that decides what ad inventory your video can hold: mid-roll ads require videos of eight minutes or longer. A seven-minute-fifty video and an eight-minute-ten video are different products on the money side.

What to look at, not just the average:

• average percentage viewed, which is comparable across different video lengths in a way the raw duration is not;
• the retention curve’s first 30 seconds. A cliff there is a promise problem — the title and thumbnail sold something the opening did not deliver;
• whether the number is flat across all videos. Identical low retention on everything is the signature of traffic that did not choose you, which is the pattern described in buying 4000 watch hours.

2. Impressions and click-through rate

Impressions are how many times YouTube offered your video to somebody. CTR is how often they took it. Together they tell you which of two very different problems you have:

• impressions holding, CTR falling — your packaging stopped working. Titles and thumbnails, nothing deeper;
• impressions falling, CTR holding — the system reduced how often it offers you. That is not a packaging problem, and chasing it with new thumbnails wastes weeks. The causes are in views dropped;
• both falling — usually seasonal or topical: the audience for the subject moved on.

The useful habit is reading these two together. Alone, each one tells a story that is often wrong.

3. Audience geography

This is the rate. Not a factor in the rate — the rate itself. Open Revenue and switch to RPM by country, and the usual discovery is that a small minority of viewers produces most of the money, because per-thousand estimates for the United States sit around $5.4 while a median market sits near $0.95.

Two consequences:

• the channel average RPM is almost meaningless. It is a weighted blend of markets that pay seven times differently;
• growth in the wrong country does not raise revenue. A hundred thousand new viewers from a cheap market can leave earnings flat, which is the single most demoralising thing in this field and the most common.

The arithmetic of what a target income costs in each market is in views needed for $1,000 a month.

4. Returning viewers

New versus returning viewers answers one question: are you building an audience or renting traffic? A channel that is almost entirely new viewers restarts from zero with every upload, regardless of subscriber count — and subscriber count is exactly what hides this, because subscribers who never come back still count as subscribers.

A healthy number differs wildly by niche, so the comparison that matters is your own channel against itself three months ago. A falling returning share while views hold is an early warning that the channel is being carried by one format that is wearing out.

5. RPM against CPM

The most common confusion in the topic, and it costs people real decisions:

• CPM is what advertisers pay per thousand ad impressions. It is a market price and it includes YouTube’s share;
• RPM is what you received per thousand video views — after YouTube’s cut, and including all the views that carried no ad at all.

So RPM is always lower than CPM, often by a lot, and the gap is not a fault. If your CPM is healthy and your RPM is poor, the money is being lost in the gap — which is the next number.

6. The share of views that carried an ad

Not every view can be monetized. Viewers with ad-free subscriptions, videos limited by advertiser suitability, regions with thin demand, and views where no ad was available all contribute views without contributing ad revenue.

If this share is low, the fix is not more views. It is finding out why: a topic that advertisers avoid, videos under eight minutes with no mid-roll inventory, a self-applied content rating, or an audience in a market with little advertiser demand. That diagnosis is also the honest answer to “why is my RPM so low” — more often the answer is here than in anything you can change about the video.

The same six on the other platform

Facebook Content Monetization pays per qualified view — Meta’s own measure of a view that counted — which collapses numbers 5 and 6 into one: the share of your views that qualified is the whole story, and there is no CPM to compare against.

What stays identical is number 3. The rate is the audience’s language and market, and on our own pages the spread between the cheapest and most expensive audience language is 53-fold: the public end runs from Russian at $0.014 per thousand to German at $0.163, and our overall measured range is $0.08 to $0.33, averaging $0.185. Format matters far less: photos $0.038, text $0.031, video $0.018–0.023.

This is also the checklist for reading somebody else’s channel before buying it — the same six numbers, in the same order, on their screen rather than yours. That version is in buying a channel or a page.

The network runs about 250 pages in 25 languages. Across the monetized pages the average is $0.024 per thousand views, and $0.08 to $0.33 per thousand qualified views — the ones Meta counts, averaging $0.185. The best single page-month was $21,216 and the best day $4,163. These are measurements, not a promise: the rate follows language, market and format. Full breakdown in RPM by country.

How we help

1. Your videos and photos go out on the network's pages and we split the revenue. The invitation, the eligible country and the payout account stop being your problem — the pages and the payouts are on us. We pick the language and the market that fit your content, and we check every file for originality risk before it is published. The rights to your material stay with you.
2. You monetize on your own page and we set it up. If you would rather own the asset and keep all of the revenue, we prepare the page, help with the payout details and hand over a plan of what to publish and when.

Write to us with a link to your channel or page: we will say which languages your material makes sense in and roughly what it would return.

Fill in the application →

Questions from the community