YouTube fake views: what the policy bans and what buyers can see
Part of YouTube brand safety
The short answer
YouTube's fake engagement policy bans anything that artificially increases views, likes, comments or other metrics, including bought views and sub4sub schemes. YouTube says artificial traffic is not counted and may lead to strikes. For buyers, inflated numbers distort a channel's apparent reach, so engagement out of line with views deserves a closer look before the channel enters a plan.
Key takeaways
- YouTube prohibits artificially inflating views, likes, comments or subscribers, and it holds creators responsible for promoters they hire.
- YouTube says it may slow, freeze or change counts while it validates them, so a dropping view count can be the system removing invalid views.
- Uneven engagement has innocent causes, including ad views that count toward public views and the start of play counting since August 24, 2026.
- Viewers report suspect videos and channels to YouTube; advertisers raise invalid activity on their own ads with Google Ads.
What does YouTube's fake engagement policy prohibit?
YouTube's fake engagement policy (opens in a new tab) says it "doesn't allow anything that artificially increases the number of views, likes, comments, or other metrics", whether through automatic systems or by serving videos to unsuspecting viewers. The policy also bans content that exists solely to incentivize engagement, and calls engagement illegitimate when it results from coercion or deception, or when its sole purpose is financial gain.
Definition
- Fake engagement
- Views, likes, comments, subscribers or other YouTube metrics produced artificially rather than by real viewers choosing to watch or respond. YouTube treats it as a policy violation whether the creator did it or someone the creator hired.
| Prohibited example | In YouTube's words | Why a buyer should care |
|---|---|---|
| Inflation services | Links to or promotes third-party services that artificially inflate metrics like views, likes, and subscribers | A channel advertising these services is selling the problem the brand wants to avoid |
| Sub4sub | Offering to subscribe to another creator's channel only if they subscribe to your channel | Subscriber totals stop meaning an interested audience |
| Bought views shown as promotion | A creator purchasing their views from a third party with the intent of promoting the service | Signals a channel built around gaming metrics |
| Dedicated channels | Channels dedicated to artificial engagement, among other examples | Such channels should never be in a plan at all |
The policy is not a ban on asking: YouTube says creators may encourage viewers to subscribe, like, share or comment. Trading subscriptions or paying for them is the line, and YouTube notes its list of examples is not complete.
What does YouTube do when it finds fake engagement?
The policy page says YouTube removes the content and emails the creator. A first violation usually brings a warning that can expire after a policy training, and three strikes within 90 days may end with the channel terminated. A single case of severe abuse, or a channel dedicated to a policy violation, can be terminated directly. The page also states that "Page traffic found to be artificial will not be counted on YouTube and can lead to strikes on your account", and that spam subscribers and terminated accounts are left out of subscriber and view totals.
Counts also move while YouTube checks them. Its page on how engagement metrics are counted (opens in a new tab) says it takes time to work out which views, likes, dislikes and subscriptions are legitimate, and that YouTube may temporarily slow down, freeze or change a metric count. It discards low quality playbacks, such as one person playing the same video on several devices or in multiple tabs. A count that stalls or drops can be validation at work.
How do fake views show up in channel statistics?
Mostly as mismatches, because buyers only see public totals. The YouTube Data API channels resource (opens in a new tab) gives a channel's total views, its subscriber count rounded down to three significant figures, whether that count is hidden, and its number of public videos. The videos resource (opens in a new tab) gives each video's views, likes and comments, while dislikes were made private on December 13, 2021 and appear only to the video owner.
| Statistic | What a buyer can read | Pattern worth a look |
|---|---|---|
| Video view count | Views per video, upload by upload | A few videos far above the rest with no visible reason, such as a trend or a press mention |
| Like count | Likes per video | Likes very low relative to views across many uploads, not just one |
| Comment count | Comments per video, and the comments themselves | Comments that repeat generic praise, arrive in bursts or ignore what the video says |
| Subscriber count | Rounded total, or hidden | Subscribers high, with typical videos drawing a tiny fraction of that number |
| Channel view total | All views across formats | A total dominated by a handful of uploads, with recent videos far lower |
None of these reads proves anything alone. What survives in public numbers can be history: a burst bought years ago that still inflates the subscriber total while new uploads draw a fraction of it. Forecasts built on that total would overstate reach.
What does uneven engagement on a channel suggest?
It suggests a question, not an answer. Several ordinary causes make engagement uneven, and some of them are created by YouTube's own counting. YouTube's metrics page says public views are counted the moment a video ad starts to play across Shorts and in-stream ads, so a video the creator promoted with ads can show many views and few likes. It also says that beginning August 24, 2026, views are counted the moment a video starts to play across all formats, and the channels resource notes this includes autoplay and hover to play.
| Pattern | Possible innocent cause | Reading that suggests inflation | Check to run |
|---|---|---|---|
| High views, very few likes | The video ran as a paid ad, and ad views count toward public views | The same ratio on every upload, including ones with no promotion | Look for an ads disclosure or a product launch; compare unpromoted uploads |
| Sudden spike on one video | A trend, an embed on a popular site, or a press mention | Spike with no outside mention and comments that do not refer to the content | Search for the video's title outside YouTube and read the comments |
| Subscribers far above typical views | An older audience that drifted away, or a format change | Subscriber jumps with no matching jump in views at the time | Compare views on uploads from the same period |
| Comments off | YouTube lists comments as unavailable on made for kids content, or the creator chose to close them | Comments off only on the videos with outsized views | Check the made for kids label and the creator's other videos |
| Views drop after a few days | YouTube removing low quality or invalid views during validation | Repeated large drops across many videos | Note the dates and treat recent counts as provisional |
Weigh patterns across ten or more recent uploads: one odd video is noise, most videos sharing the same odd shape is a pattern. See the channel vetting checklist for the wider review.
Can a channel with bought views still be in a plan?
Sometimes. It depends on whether the inflation is current and whether the plan's numbers can be rebuilt from real views. A channel that promotes inflation services, trades subscriptions or shows inflated numbers on its newest uploads stays out; one whose only problem is an old burst can be judged on its recent work.
Deciding on a channel with signs of inflated metrics
- Date the inflation. Identify which uploads carry the suspect numbers and when they were published.
- Check the current uploads. If the latest ten videos show natural ratios, the problem may be history rather than practice.
- Look for promotion of inflation. Any link to or plug for a view or subscriber service removes the channel, since YouTube names it as prohibited.
- Rebuild the numbers. Base the channel's forecast on median views of recent uploads, never on subscriber totals or channel lifetime views.
- Record the decision. Note the suspect videos, the dates and the reasoning, so a later reviewer sees why the channel stayed or left.
How do you report suspected fake engagement?
Through two separate routes, depending on what is suspected. YouTube's reporting help (opens in a new tab) says a signed in user can report a video from More, then Report, choosing the reason that best fits, or report a channel from the channel description by selecting more, then Report user. Reporting is anonymous, reported content is not taken down automatically, and it is reviewed against YouTube's guidelines.
Suspicion about the brand's own ad delivery is a different matter. Google's invalid traffic page (opens in a new tab) says invalid traffic found before the end of a billing cycle is removed from metrics and billing, that later findings may lead to credits shown on invoices, and that an advertiser can request an investigation for activity within the last 60 days through the Click Quality Form. The invalid traffic guide explains the wider categories.
What to keep before reporting or escalating
- The channel ID and the video IDs concerned, not only the handle.
- Screenshots or exported counts with the date they were taken, since YouTube may change counts while validating.
- A short description of the pattern, such as likes and comments far below views on most recent uploads.
- For the brand's own ads, the Customer ID, date range and campaign and ad group names Google asks for.
- A note of whether the channel was removed from the plan, and by whom.
What does a fake views check look like on a real brief?
Fictional example
A luggage brand screens travel channels in Wisconsin
An unnamed maker of carry-on luggage and packing cubes plans YouTube ads in Milwaukee and Madison, in English, beside travel tips, packing guides and weekend trip vlogs.
One candidate travel channel shows a large subscriber total, yet its last ten uploads draw a small fraction of it, and two older videos hold most of the channel's lifetime views with almost no comments. Its recent uploads look natural. The planner keeps the channel but rebuilds its forecast on recent median views, noting the old videos as history.
A second channel links to a subscriber growth service in every description. It leaves the plan, and the planner reports the channel to YouTube from its description, keeping dated screenshots in the plan's notes.
How does PlacementGraph catch unusual views per subscriber?
Frequently asked questions
Are autoplay views on YouTube fake views?
No. YouTube says that beginning August 24, 2026, views are counted the moment a video starts to play across all formats, and the Data API notes this includes autoplay and hover to play. Higher totals after that date reflect a counting change, not inflation.
Do sub4sub groups break YouTube's rules?
Yes. YouTube's fake engagement policy lists offering to subscribe to another creator's channel only if they subscribe back as a prohibited example. Subscribers gained that way do not represent an interested audience.
Can a creator lose a video over views a promoter bought?
Yes. The policy says content can be taken down whether the action was taken by the creator or by someone the creator hired, and warns that a promoter's decisions may affect the channel. Repeated violations lead to strikes, and three strikes within 90 days may end in termination.
Does a hidden subscriber count mean a channel is hiding fake subscribers?
No. The Data API simply reports whether a channel's subscriber count is publicly visible, and a creator can hide the count for reasons that have nothing to do with fraud. Views, likes and comments on recent uploads stay public and are the better evidence.
Sources
- Fake engagement policy (opens in a new tab)YouTube Help
- How engagement metrics are counted (opens in a new tab)YouTube Help
- Channels resource (YouTube Data API) (opens in a new tab)Google for Developers
- Videos resource (YouTube Data API) (opens in a new tab)Google for Developers
- Report inappropriate videos, channels, and other content on YouTube (opens in a new tab)YouTube Help
- About invalid traffic (opens in a new tab)Google Ads Help
- Watching "made for kids" content (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.