Ask new creators about YouTube monetization and they'll quote the subscriber requirement. Ask monetized creators which one almost stopped them and most say the other: the watch hours. Subscribers accumulate forever. Watch hours are measured over a trailing year, which is the part almost nobody explains properly.
There's now a second part almost nobody explains: the number itself moves. Depending on when you cross, the bar is either 4,000 hours or 8,000. The mechanics are identical either way, which is why this guide is built around them and not around one figure.
Warning
Work out which bar is yours before you plan anything. Get accepted before February 1, 2027 and the requirement is 4,000 qualified watch hours in the last 12 months. Get there after that date and it is 8,000 over 365 days, or 20 million qualified Shorts views over 90 days, alongside the same 1,000 subscribers. Channels already accepted keep the old bar. Our watch hours calculator picks the bar for you: it projects your crossing month, then measures you against the threshold in force that month. Full breakdown of the announcement in what changes in 2027.
The requirement, exactly
To join the YouTube Partner Program at the full ad-revenue tier, you need 1,000 subscribers plus one of the following:
| Now, through Jan 31, 2027 | From Feb 1, 2027 | |
|---|---|---|
| Qualified watch hours | 4,000 in the last 12 months | 8,000 in the last 365 days |
| or qualified Shorts views | 10 million in 90 days | 20 million in 90 days |
It's either/or, not both, no matter what a YouTube guru told you. "Qualified watch hours" is YouTube's current wording, replacing the "valid public watch hours" still quoted everywhere. You also need an eligible country, no active Community Guidelines strikes, 2-Step Verification and advanced features on, and an AdSense for YouTube account linked or ready to create. Those rarely stop anyone. The hours do.
Put the watch-hours column in daily terms: 4,000 hours across 365 days is about 11 hours of watch time a day, roughly 164 views a day at a 4-minute average view duration. 8,000 hours is about 22 hours a day, or 329 views. The Shorts alternative works out around 676 times those view counts, and that ratio doesn't budge, because both bars doubled together.
Check Which Requirements You Already MeetThe 500-subscriber tier does not move
Under the headline requirement is a gentler door, and it's the one piece February 2027 leaves alone: 500 subscribers, 3 valid public uploads in the last 90 days, and either 3,000 qualified watch hours in 12 months or 3 million qualified Shorts views in 90 days. Documented as unchanged, not merely unmentioned. YouTube's words: "There are no changes to the eligibility requirements for fan funding, YouTube Creator Partnerships, or YouTube Shopping."
No ad revenue or Premium here, but you get the fan-funding stack: memberships, Super Chat and Super Stickers, Super Thanks, jewels and gifts, and Shopping. Aim here first, and note it matters more than it did a year ago. 3,000 hours is 75% of the old bar and 37.5% of the new one, so after the cutover it's the milestone that still arrives on a human timescale.
Warning
The 500 tier has a catch the 1,000 tier doesn't: 3 valid public uploads in the last 90 days. If you burned through your watch hours a year ago and then went quiet, you can hold 3,000 hours and still be ineligible. Post something.
What counts, and what quietly doesn't
"Qualified" is doing a lot of work in that phrase. YouTube publishes a positive definition and an exclusion list, and the exclusions catch people constantly:
| Source of watch time | Counts? | Per YouTube? |
|---|---|---|
| Public long-form videos | Yes | Stated |
| Public live streams converted to VOD and kept up | Yes | Stated |
| Organic follow-on views after promoted content | Yes | Stated |
| Shorts | No | Stated |
| Private, unlisted, or deleted videos | No | Stated |
| Views from ad campaigns you paid for | No | Stated |
| Streams left unlisted, deleted, or never converted to VOD | No | Stated |
| Premieres | Probably yes | Not addressed |
| Your videos embedded on other sites | Probably yes | Not addressed |
That last column matters, and most articles on this topic won't give it to you. The first seven rows are YouTube's own words. The last two are our read: Premieres and embedded playback are ordinary public long-form watch time and nothing on the exclusion list touches them, so they should count. YouTube has never said so directly, so plan conservatively around either.
Three rows deserve a closer look.
Shorts don't count. At all. The exclusion list names "YouTube Shorts" flatly, and YouTube adds that "any qualified watch hours from Shorts views in the Shorts Feed don't count towards the 4,000 qualified watch hours threshold." A Short with a million views builds the Shorts-views path and contributes exactly 0 watch hours. If your channel is mostly Shorts, you're on the other path whether you meant to be or not.
Deleted videos probably take their hours with them. YouTube's exclusion list names private, unlisted and deleted videos, but it never says whether the check looks at a video's status today or its status when each view happened. The plain reading, and what creators consistently report, is today's status: delete an old upload you're embarrassed by and the hours it earned inside the window stop counting. Private most likely does the same. YouTube hasn't confirmed the mechanism, so leave bad-but-harmless videos up until you're monetized.
Paid views don't work, but the follow-on exception is real. Ad-campaign traffic is excluded by policy, and purchased watch-hour "services" get filtered as fake engagement, which can cost the channel rather than just the hours. YouTube then carves out an exception almost nobody quotes: Follow-on Views, "organic views you get when someone watches your videos after seeing your promoted content," count. The ad view itself never does.
Hours leave the window, they don't pile up
YouTube words the requirement as hours "in the last 12 months," elsewhere "in the last 365 days." It has never published the underlying mechanics and has never used the word "expire," so treat what follows as the plain reading rather than documented gospel: the window slides, and hours earned more than a year ago sit outside it. Nothing resets on January 1st, and nothing resets when you apply.
Read that way, it has a brutal implication for slow channels. Stay flat below about 334 watch hours a month and you never reach 4,000; below about 667 a month you never reach 8,000, no matter how long you upload. Old hours fall out the back as fast as new ones arrive at the front. Both floors are just the bars divided by twelve.
Most watch-hour calculators ignore this and add your hours up forever, which tells slow channels a comforting lie. Ours models the rolling window: plug in your hours, monthly views and average view duration, and it projects the month your rolling total crosses the bar in force that month, or tells you your pace never gets there and how many extra views a month would.
Project Your Crossing Month Against Both BarsThe math: minutes divided by average view duration
4,000 hours is 240,000 minutes. 8,000 hours is 480,000. Divide by your average view duration and you get the views you need across a rolling year:
| Average view duration | Views for 4,000 hrs | Per month | Views for 8,000 hrs | Per month |
|---|---|---|---|---|
| 1 minute | 240,000 | 20,000 | 480,000 | 40,000 |
| 2 minutes | 120,000 | 10,000 | 240,000 | 20,000 |
| 4 minutes | 60,000 | 5,000 | 120,000 | 10,000 |
| 5 minutes | 48,000 | 4,000 | 96,000 | 8,000 |
| 8 minutes | 30,000 | 2,500 | 60,000 | 5,000 |
| 10 minutes | 24,000 | 2,000 | 48,000 | 4,000 |
Redo any row yourself: hours × 60 ÷ average view duration in minutes = views per year, then ÷ 12 for the monthly figure. Average view duration lives in YouTube Studio under Analytics, then Engagement, shown as m:ss, so 3:30 is 3.5.
That table is why average view duration is the highest-leverage number on your channel. Hold viewers for 8 minutes and you need one-eighth the audience of someone holding them for 1 minute, at either bar.
For calibration, Retention Rabbit's benchmark report on more than 10,000 videos put average retention at 23.7%, with the 5-to-10-minute bucket best at 31.5%. Vendor study, not platform data, and its URL now 404s, so treat it as directional. A 10-minute video held at 31.5% is a 3.15-minute average view duration: about 76,000 views a year for 4,000 hours, 152,000 for 8,000. Hold it at 40 to 50% and you're at 4 to 5 minutes, so 48,000 to 60,000, or 96,000 to 120,000.
What your current pace gets you
Take the hours you earn per month now and the rolling window does the rest, assuming a steady pace and a standing start at zero banked hours:
| Watch hours per month | Rolling total after 12 flat months | Crosses 4,000 | Crosses 8,000 |
|---|---|---|---|
| 100 | 1,200 | Never | Never |
| 200 | 2,400 | Never | Never |
| 334 | 4,008 | Month 12, barely | Never |
| 400 | 4,800 | Month 10 | Never |
| 600 | 7,200 | Month 7 | Never |
| 667 | 8,004 | Month 6 | Month 12, barely |
| 1,000 | 12,000 | Month 4 | Month 8 |
| 1,500 | 18,000 | Month 3 | Month 6 |
Sit with the middle of that table. A channel banking 400 hours a month is not failing; it clears the old bar inside a year and never clears the new one. That's the whole policy change in one row. For the rows that never arrive, the fix isn't another year of uploading. It's changing what you upload: longer videos, better retention, or search topics that keep pulling views for months. Usually all three.
"I'm at 3,500 hours on January 31, 2027"
This is the scenario worth working through, because the panic version of it is wrong.
Your 3,500 hours do not vanish on February 1. They were earned inside the trailing 365 days and they stay inside it until they age out on their own schedule, exactly as they would have anyway. Nothing about the cutover reaches back and deletes anything. Only the height of the line changes: you were 500 hours from the target, now you're 4,500 from it.
That's a real setback, but a different kind than losing progress. Here's the cost at a few paces, from 3,500 banked hours earned evenly across the previous year, the assumption our calculator makes and discloses:
| Watch hours per month from here | You cross 8,000 |
|---|---|
| 400 | Never at this pace. Settles at 4,800 |
| 667 | Month 12, by four hours |
| 1,000 | Month 7 |
| 1,500 | Month 4 |
The 667 row is the honest floor: hit exactly the pace the new bar demands and you arrive twelve months later with four hours to spare. Anything less and the window swallows your progress. It's also the clearest argument for the 500-subscriber tier, since 3,500 hours already clears the 3,000 needed there and fan funding doesn't care what happens in February.
One case genuinely has no published answer: creators who applied in January and are still under review when the date passes. YouTube documents that your counts dropping after you apply doesn't matter, a different situation from the bar rising while you wait, and it says nothing about the second. It sits among the open questions in what changes in 2027.
Run Your Own Hours Against the CutoverChecking your number in Studio
YouTube Studio, Earn tab. That's the only number that matters, because it's the one YouTube measures against the threshold. There's also a "Get notified" option there that emails you once you hit an eligibility threshold.
The Analytics tab will happily show you a watch-time figure too, and this is where people confuse themselves. Analytics reports whatever date range you selected, and it counts watch time the threshold doesn't, Shorts being the obvious one, so the two routinely disagree. That last part is observed practice rather than anything YouTube documents; creators hit the gap constantly, but no support page spells it out. Go with the Earn tab.
What actually moves watch hours
Ranked by leverage, from the arithmetic above rather than vibes.
1. Live streaming, the multiplier most small channels skip
Watch hours from a public stream are concurrent viewers times stream length, and the archive keeps earning afterward as a normal video. A 2-hour stream with 30 people watching banks 60 hours in an evening; the same 60 hours needs 720 views of a 10-minute video held at 50%. If your content works live at all, streaming is the fastest legitimate accumulator there is. One catch, from the exclusion list: convert the stream to VOD and leave it public.
2. Retention beats length
A 10-minute video watched at 80% beats a 20-minute video watched at 30%, and padding runtime actively hurts for a reason that needs no study behind it: retention is a percentage, so every minute you haven't earned drags the average down. Cut the channel-intro throat-clearing. Make the video as long as the content deserves, then work on holding attention rather than adding minutes.
3. Evergreen search content compounds inside the window
The rolling window punishes spiky content. A trend video earns hours for three weeks, then ages out a year later, right when you need it. A how-to video pulling 1,500 search views a month at a 4-minute duration adds 100 hours to your rolling total every month, indefinitely. Five of those is 500 hours a month, most of the 667 floor from search alone.
4. Playlists and session chaining
Watch hours count wherever your public videos get watched, so the cheapest extra hours come from viewers already watching. Order playlists so videos hand off naturally, point end screens at the logical next video, and say the handoff out loud. Chained views retain worse than the one that pulled someone in, so treat it as an upper bound. But a 20% bump in videos-per-session is close to a 20% bump in hours from the same audience.
What doesn't work
Uploading daily instead of uploading well. Frequency only helps if each upload holds attention, and a channel sprinting to publish daily almost never clears 40% retention. Ten videos a month at a 2-minute average view duration and 200 views each is 67 hours. Two videos a month at 5 minutes and 1,200 views each is 200 hours.
Sub4sub and watch-time pods. Reciprocal-viewing groups produce exactly the signature YouTube's fake-engagement systems look for. The hours get filtered, and the attempt is a policy violation.
Community posts, polls, and Shorts as a warm-up. Fine for the audience, worth zero watch hours. And be careful with a piece of advice that circulates here: Community posts count as channel activity for the inactivity rule that applies after you're monetized, but they are not "uploads" for the 500-sub tier's 3-in-90-days requirement. Two rules, two clocks.
Re-uploading other people's content with light edits. Compilations, reaction cuts with no commentary, and clip channels can accumulate hours perfectly well and fail review anyway.
You still have to pass review
Hitting the bar doesn't monetize you; it lets you apply. You have to finish the signup steps first, including signing the contract and linking an AdSense for YouTube account, before review starts. Then automated systems and human reviewers check the channel as a whole, typically about a month. YouTube doesn't rank rejection reasons publicly, but two of its named content policies cover most of what trips creators up:
- Reused content: other people's material with minimal changes and no transformative value.
- Inauthentic content: mass-produced, generic, repetitive uploads. YouTube renamed its "repetitious content" policy to this on July 15, 2025. Enforcement against AI-slop channels has visibly tightened since. In January 2026, The Verge reported that YouTube confirmed removing 11 channels flagged in a Kapwing analysis for violating its spam, deceptive practices and scams policies, with one more suspended from the Partner Program; most of the other flagged channels were taken down by their own owners. The 11 is the only YouTube-confirmed number, and it is a Community Guidelines action rather than a YPP review outcome, but it tells you which way the wind is blowing.
If you're rejected you can appeal within 21 days or reapply after 30 days, 90 days on a repeat rejection. Your hours don't disappear while you wait, though the oldest keep aging out.
One thing folklore gets wrong: once you're in, dipping back under the bar does not automatically remove you. The stated removal triggers are policy violations and inactivity, not a threshold re-check.
That said, "inactivity" is itself redefined on February 1, 2027, and this part has barely been covered anywhere. From that date a channel counts as active with 1,000 qualified watch hours in the past 365 days, or 1 million qualified Shorts views in 90 days, or 2 long-form videos or 5 Shorts every 90 days. Fall below all three and you get an extended 90-day window to meet one of the first two. So a watch-hours number is now attached to staying in, not just getting in. It's 1,000, an eighth of the new entry bar and a quarter of the old one.
Frequently asked questions
Do my existing watch hours still count after February 1, 2027?
Yes. Nothing is reset or wiped on the cutover date. Hours earned inside the trailing 365 days keep counting and age out on their own schedule. What changes is the target: 8,000 rather than 4,000, for anyone not yet accepted into the program. A channel at 5,000 hours on January 31 is over the line that day and 3,000 short the next, holding the same hours.
If I qualify in January 2027, do I keep the 4,000 bar?
If you're accepted before the cutover, yes. YouTube is explicit: "If you are already in YPP, your status is not impacted by this update." Being accepted is what grandfathers you, not being close. The unresolved case is the one in between, where you applied in January and review is still running on February 1. YouTube says a count dropping after you apply doesn't matter, but that's a different situation from the bar rising mid-review, and it has published nothing about the second. Reviews take about a month, so apply the moment you're eligible.
Can I buy watch hours to get monetized?
You can spend money, but you can't buy monetization. Purchased watch time gets filtered by YouTube's fake-engagement systems before it counts, and the attempt is a policy violation that can get your channel removed or terminated. Paid ad-campaign views are excluded too, so "just run ads" doesn't work. The one narrow exception is organic follow-on views: someone sees your promoted video, watches your others on their own, and that watch time counts.
What happens if my watch hours drop below the bar after I'm monetized?
Nothing automatic. Falling back under 4,000 or 8,000 is not a removal trigger; the published triggers are policy violations and inactivity. Claims that YouTube revokes monetization the moment your total dips are folklore. The nuance: from February 1, 2027, "active" gets its own numeric definition, 1,000 qualified watch hours in 365 days or 1 million Shorts views in 90 days or 2 long-form videos or 5 Shorts every 90 days. Drop below all three and you get a 90-day window to recover by clearing one of the first two; the upload-count option carries no grace period of its own.
How long does it take to get 1,000 subscribers?
Nobody publishes a credible watch-hours timeline, so the best public data covers the subscriber half. vidIQ analyzed 9,414 channels created after 2023 that crossed 1,000 subscribers and found a median of roughly 16 months, the fastest quarter under 8.5 months, the slowest at two years or more, and 64% taking over a year. That's vidIQ's published sample, not platform data, and vidIQ is an affiliate partner of this site, so treat it as a data point with a stated sample size rather than authority. It's also what makes the 2027 change concrete: on that timeline, most people starting today get judged against 8,000 hours. For what the subscriber requirement itself involves, and why it is the easier half, see how many subscribers you need to make money.
Sources: YouTube Help, all re-fetched as raw HTML on August 22, 2026. YPP overview and eligibility: both thresholds, the qualified-hours definition and exclusion list, "Get notified", review timing, appeal and reapply windows. Changes to the YPP: the 8,000 / 20M entry change for new creators, the unchanged fan-funding tier, the 2027 channel-activity redefinition. Expanded YPP tier. Check your YPP eligibility: the Follow-on Views exception. Channel monetization policies: the July 15, 2025 rename to inauthentic content. Retention benchmarks are Retention Rabbit's 2025 report, cited from an archived copy because its URL now 404s; subscriber timelines from vidIQ, an affiliate partner; the January 2026 enforcement figures are The Verge's, linked inline.
