Which YouTube channel metrics should advertisers read?
Part of YouTube channel planning
The short answer
Advertisers should read four public channel metrics before buying: median views on recent uploads, views per subscriber, upload pace and the age of the newest video. Each comes from public counts in the YouTube Data API. Read them together and by format, because lifetime views or subscribers describe the channel's past more than its current audience.
Key takeaways
- Median views on the last ten to twenty uploads of one format describe a channel's current pull better than its lifetime view total.
- Views per subscriber is a rough activity ratio, since YouTube says subscriber count is not the most accurate way to estimate an active audience.
- Upload pace matters for flight coverage: a channel that rarely publishes gives an ad few new videos to appear beside during the flight.
- Uneven engagement is read as a distribution, so one breakout video and a steady body of work are told apart before anyone suspects manipulation.
Which public channel metrics should advertisers check first?
Four derived numbers anchor a channel decision, and all of them can be built from the public counts in the YouTube Data API (opens in a new tab). The channels resource gives a channel's total views, its subscriber count, whether that count is hidden and its number of public videos. The videos resource (opens in a new tab) adds each video's views, likes, comments, publication time and duration. The table shows what each derived metric tells a buyer and where it stops being useful.
| Metric | Built from | What it tells a buyer | Where it misleads |
|---|---|---|---|
| Median views per recent upload | statistics.viewCount on the last 10 to 20 videos of one format | The views a typical new video draws now, the closest public proxy for what a placement can reach | Young videos are still accruing views; one format can hide another's pull |
| Views per subscriber | Median recent views divided by statistics.subscriberCount | Whether the subscriber base still watches, or the count is history | Rounded or hidden counts; Shorts and long form react differently |
| Upload pace | Gaps between snippet.publishedAt dates | How many new videos the flight can expect | Batch publishing and seasonal channels break simple averages |
| Recency | Date of the newest public upload | Whether the channel is active at all | A recent upload says nothing about its topic or quality |
Lifetime totals are left out of the first column on purpose. The channels resource defines viewCount as the sum of views of all the channel's videos in all formats, so one old viral clip can give a quiet channel the same total as one drawing steady views every week.
Which channel metrics can a buyer check without channel access?
Definition
- Public channel metrics
- Counts YouTube publishes for anyone to read, on the channel page or through the YouTube Data API, such as views, subscribers, likes, comments and upload dates. Creator reports, such as monthly audience, sit inside YouTube Studio and reach a buyer only if the creator shares them.
The split matters because the two groups answer different questions. Public counts show scale and activity. YouTube's audience report guide (opens in a new tab) lists the reports only the creator sees, including monthly audience, which it defines as "the total unique viewers who watched your content in the last 28 days", watch time from subscribers, and top geographies, which shows "which geographies have the most watch time for your channel."
| Metric | Public | Creator report only |
|---|---|---|
| Views per video and channel total | Yes | Realtime estimates also appear in YouTube Analytics |
| Subscriber count | Yes, rounded down to three significant figures, or hidden | Watch time from subscribers, as a share of watch time |
| Likes and comments | Yes, per video | Dislikes, private since December 13, 2021 and returned only to the video owner |
| Upload dates and durations | Yes | Not needed |
| Monthly audience, unique viewers | No | Yes, last 28 days |
| Top geographies and age ranges | No | Yes, measured by watch time |
What does views per subscriber reveal about a YouTube channel?
Views per subscriber shows whether the people who once subscribed still watch. YouTube's own guide to new, casual and regular viewers (opens in a new tab) says "Subscriber count is not the most accurate way to estimate the size of your active audience" and that "It's also common for viewers to be subscribed to channels they no longer watch at all." A ratio built on recent median views tests how much of that base is still active.
Computing views per subscriber in a way two planners would repeat
- Pick one format. Use the last ten to twenty long form uploads, or the last ten to twenty Shorts, never a mix, because the channel total combines all formats.
- Drop videos younger than about two weeks. Their counts are still climbing, and YouTube's engagement metrics page (opens in a new tab) says metrics may take time to appear in the first few hours and are constantly confirmed and adjusted.
- Take the median, not the average. One breakout video lifts an average far more than it lifts a median.
- Divide by the public subscriber count. Note that the API rounds it down to three significant figures, so the ratio is approximate by construction.
- Record the date and the sample. The ratio changes as videos age, so a number without its date cannot be compared later.
Read the result as a direction, not a grade. A low ratio on a large channel suggests a past audience that drifted, which matters for any forecast built on the subscriber total. A high ratio on a small channel suggests videos that travel beyond subscribers. Neither reading says anything about the quality of the views; the fake views guide covers that.
How should upload pace affect a YouTube channel decision?
Upload pace is a planning metric before it is a quality metric. A channel placement covers the channel's videos, so a channel that publishes twice a month gives a four week flight two new videos plus its back catalog. YouTube's performance troubleshooting page (opens in a new tab) adds a caution: it says YouTube studied thousands of channels that took a break "and found no correlation between break length and changes in views." A pause alone does not show the audience left.
YouTube's channel health tips (opens in a new tab) say the opposite side as well: "publishing new videos is correlated with a growth in views." Taken together, the two pages support a practical rule. Treat pace as a question of how much fresh inventory a flight will see, and treat a sudden change in pace as a reason to read the channel's recent uploads, not as a verdict.
| Pattern | What it suggests | What to check |
|---|---|---|
| Steady weekly or faster | Fresh videos throughout the flight | Whether the topic of recent uploads still fits the brief |
| Batches, then silence | A seasonal or series channel | Whether the next batch falls inside the flight dates |
| No upload in several months | Placements rely on the back catalog | Whether old videos still draw views at all |
| Sharp rise to many uploads a day | A change of format, team or owner | Reused or templated uploads, read in the vetting pass |
How do you judge suspiciously uneven channel engagement?
Start by describing the shape of the numbers before deciding what caused them. Uneven engagement can have an ordinary cause: YouTube's troubleshooting page describes a video that "went viral" while "those viewers didn't return to watch more." The shape tells a buyer which question to ask next, and the policy side of inflated counts belongs to the vetting and fake views reviews.
Plan on the median when a single video carries the views and the rest of the sample is consistent: the breakout is real but not repeatable, and a forecast should ignore it. Ask for a closer review when the ratios disagree across the sample, for example when likes and comments per view drop sharply only on the videos with outsized views. YouTube's channel health tips say "there's no universal 'good' benchmark for performance metrics", so compare a channel with its own history and with channels of the same format, never with a fixed industry ratio.
What does a channel metrics review look like on a real brief?
Fictional example
Model railroad channels for a hobby retailer in Ohio
An unnamed model railroad and hobby supply retailer with stores in Columbus and Cincinnati wants YouTube ads beside layout building, scenery and train operation videos for viewers in Ohio. The planner shortlists five channels and computes the four metrics from public data for each.
A large channel shows a high lifetime view total, but its median views on recent long form uploads are a small fraction of its subscriber count, and its biggest video is nine years old. The planner keeps it, with forecasts on recent medians. A smaller layout channel publishes in batches each winter, and the November flight catches its new series. A third channel's views are carried by one breakout Short; the planner plans on its long form median and asks the creator for the top geographies report to confirm an Ohio audience.
How does PlacementGraph read channel metrics for a plan?
What should a channel metrics record contain before the plan is approved?
Channel metrics record
- The sample is named: format, number of uploads and their date range.
- Videos younger than about two weeks were left out of the median.
- Median views, views per subscriber, upload pace and newest upload date are each written down with the date they were read.
- The subscriber count is marked as rounded, or as hidden when the creator hides it.
- Any breakout video is named, and the forecast uses the median without it.
- Creator reports used in the decision show their date range and metric definition.
A record like this lets another planner repeat the read later and see what changed. For the wider decision, start from the YouTube channel planning guide and the channel vetting checklist.
Frequently asked questions
Is a channel's lifetime view total a good measure of its size?
Only as background. The YouTube Data API defines it as the sum of views of all the channel's videos in all formats, so it mixes Shorts, long form and live, old and new. Median views on recent uploads of one format describe what a placement can reach today far better than the lifetime total does.
Why do brand new videos show lower view counts than expected?
YouTube says metrics may take time to appear in the first few hours after publishing and that it constantly confirms and adjusts engagement events. A video a few days old is still accruing views, so leave it out of any median used for planning.
Should Shorts views be compared with long form views on the same channel?
No. Compute each format's metrics separately. Since August 24, 2026, YouTube counts a view the moment a video starts to play across all formats, and Shorts and long form reach viewers through different feeds, so their counts are not interchangeable. Plan each format on its own median.
Do likes and comments matter for an ad placement decision?
They help as ratios within one channel, not as goals. A steady likes per view ratio across recent uploads suggests a consistent audience; a sudden break in the ratio on a few videos is worth a look. Raw like totals add little, because they scale with views.
Sources
- Channels resource (opens in a new tab)YouTube Data API, Google for Developers
- Videos resource (opens in a new tab)YouTube Data API, Google for Developers
- Understand new, casual, and regular viewers (opens in a new tab)YouTube Help
- YouTube performance FAQ and troubleshooting (opens in a new tab)YouTube Help
- Tips on channel health (opens in a new tab)YouTube Help
- Understand your YouTube audience (opens in a new tab)YouTube Help
- How engagement metrics are counted (opens in a new tab)YouTube Help
Researched and written by PlacementGraph Research. Read how we research, source and update these pages. Corrections: diana@placementgraph.com.