Key takeaways
- The fourth quarter is real, large and measurable — Alphabet's reported YouTube ad revenue went from $10.26bn in the third quarter of 2025 to $11.38bn in the fourth, then fell to $9.88bn in the first quarter of 2026. The same shape appeared a year earlier.
- The collapse in January is an advertising event, not an audience event. Nielsen recorded total US TV viewing at a twelve-month high in January 2026, up 3.7% on December. The viewers stay; the bidders leave.
- Seasonal demand has to be published into on the way up. Google's own holiday research found 83% of early-October purchases were researched in advance, up from 77% a year before — the deciding happens weeks before the buying.
- Q4 is the one quarter where every competitor's packaging converges on the same red, green and gold. Signalling the season should cost you a corner of the frame; spending the whole frame on it makes you invisible.
- The money that is not in the ad auction — sponsorships, affiliate commission, memberships — is negotiated in September and October, not in December.
- Q5, roughly 26 December to mid-January, is the cheapest attention of the year: high intent, low advertiser competition, new devices in people's hands, and almost nobody publishing against you.
It is the middle of September, which means the next four months of your channel are already mostly decided. Not by what you upload in November — by what you start filming now, what you agree with a sponsor in the next three weeks, and whether the video that is going to carry your December was published before the demand for it peaked or after.
Most creators meet the fourth quarter twice. The first time is pleasant: revenue per thousand views climbs through November, December looks like the channel finally worked, and Studio flatters everything about the strategy. The second is not: same channel, same videos, same audience, and a January revenue line that reads like a punishment. Then comes the mistake this article exists to prevent — changing the channel in response to the calendar.
What follows is the quarter worked out properly: how big the swing is in numbers that are reported rather than estimated, when each kind of seasonal content has to be published, how to package for a feed where everyone has suddenly gone red and green, what to do with the strange fortnight after Christmas, and how to read January without panicking.
What the fourth quarter is actually worth
Almost every figure you will read about Q4 uplift comes from a tracker reconstructing CPMs out of creator screenshots. Those tables disagree with each other by several times, for the reasons set out in the piece on what YouTube pays per thousand views. There is, however, one set of numbers about YouTube's advertising that is audited and published quarterly: Alphabet reports YouTube ads as its own line item.
| Quarter | YouTube ad revenue | Change on previous quarter | Context |
|---|---|---|---|
| Q4 2024 | $10.47bn | — | First quarter above $10bn; boosted by US election spending |
| Q1 2025 | $8.93bn | About −15% | Up around 10% year on year even so |
| Q3 2025 | $10.26bn | — | Up about 15% year on year |
| Q4 2025 | $11.38bn | About +11% | A record quarter; YouTube's full-year revenue including subscriptions passed $60bn |
| Q1 2026 | $9.88bn | About −13% | Up around 11% year on year |
| Q2 2026 | $11.06bn | About +12% | Up about 13% year on year, credited largely to live sports |
Read the third column. Year on year, every quarter grew. Quarter on quarter, the same thing happens each time the calendar rolls over: the holiday quarter is the biggest of its year, and the one that follows is roughly 13 to 15% smaller in absolute terms despite the platform being larger than it was twelve months earlier. That is the seasonality, measured on the only ledger nobody reconstructed.
Two honest caveats. The Q4 2024 record was inflated by US election advertising, as Variety's coverage of that quarter noted, so the drop into Q1 2025 is partly a political calendar rather than a retail one. And the second quarter of 2026 came close to the previous December on the strength of live sport, which suggests YouTube's revenue year is becoming less lopsided. Neither changes the planning conclusion: December pays, January does not, and a gap in the low teens of percent at platform scale arrives much sharper on a single channel with a narrower advertiser base.
What you should not do is attach a multiple to it. Published trackers will tell you December pays anywhere from a third more than January to three times as much, depending on which one you open. The direction is not in dispute. The magnitude on your channel depends on your audience's country, your niche's advertiser mix, your video lengths and your format split, and the only reliable source for it is your own Studio revenue tab from last year.
Why the money moves and the audience does not
The mechanism is unglamorous. Advertising budgets are annual and are spent on a calendar. Retailers concentrate their spending where the buying is: the US National Retail Federation reported that 2025 holiday sales between 1 November and 31 December grew 4.1% to between $1.01 and $1.02 trillion, the first time the period had passed a trillion dollars, with a record 202.9 million people shopping across the five days from Thanksgiving to Cyber Monday. Brand teams with unspent budget flush it before the year closes. All of that money arrives in the same auction, at the same time, chasing the same impressions — and an auction with more bidders clears higher.
On 1 January the budgets reset to zero and the new ones take weeks to be approved and released. Nothing about your videos changed. The number of people willing to pay to interrupt them did.
Here is the part most creators never check. Nielsen's The Gauge reported that total US television viewing hit a twelve-month high in January 2026, rising 3.7% on December, with streaming holding around 47% of all TV time — and YouTube, at roughly an eighth of the whole, the largest single distributor in the measurement. December 2025 had itself set streaming records. Cold weather, dark evenings and new devices keep people watching. Demand in January is not weak; it is at its annual peak.
That one fact reorganises the quarter. If viewing is high and money is low in January, January is the cheapest month to learn something and the worst month to draw conclusions about the channel's health. And if money is high in December, December is for publishing what you have already proven works, not for a format you have never tested.
RPM is a price, not a grade
Revenue per thousand views is the output of an auction you do not participate in. When it climbs sharply in the first week of December, your videos did not improve that week; when it falls back in January, they did not get worse. The metrics that actually grade your work — click-through rate, average view duration, returning viewers — are the ones that stay comparable across the calendar. Judge the channel on those and use RPM for cash-flow planning only.
The Q4 calendar, working backwards from December
Seasonal content fails far more often from timing than from quality. A gift guide published on 12 December is competing against videos that have been accumulating watch time and ranking signals since October, in front of viewers who have already decided what they are buying. Google's holiday research is blunt about how early that decision forms: 83% of purchases made in early October 2025 had been researched beforehand, up from 77% a year earlier, and by mid-October 40% of shoppers said they were absolutely certain what they were going to buy, against 32% the year before.
| Window | What the audience is doing | What to publish | Packaging emphasis |
|---|---|---|---|
| Mid-September to mid-October | Early research; deciding rather than buying | Comparison and "which one should I buy" videos, buyer's guides, category explainers | Product legible at small size; no seasonal styling yet |
| Mid-October to mid-November | Shortlisting; watching reviews of two or three options | Gift guides, deep reviews, "best X for Y" lists | First seasonal cue appears; keep channel identity dominant |
| Black Friday week | Buying, fast, on phones | Deal roundups, live reactions, what is actually worth it | Urgency and specificity: a number, a date, a verdict |
| 1 to 20 December | Last-minute gifts, delivery cut-offs, seasonal viewing | Last-minute guides, seasonal formats, your strongest evergreen swing | Full seasonal palette acceptable — differentiate on subject, not colour |
| 26 December to mid-January | New devices, returns, self-gifting, resolutions | Setup guides, "I got X, now what", best-of, year-ahead, first-week-of-a-habit content | Back to channel identity; seasonal styling now dates the file |
Publish into the rising line, not the peak
Set Google Trends to YouTube Search and compare a seasonal query over five years. The repeating annual wave tells you when interest starts moving, and the answer for most gift-related queries is October, with the peak in December. The publishing rule follows from how a video is tested: the first days after upload decide how far it travels, so you want that test to happen while demand is still growing and competing supply is thin. Arriving at the peak means being judged against videos that have already banked their watch time. Four to six weeks before the crest is the target, which for Christmas gifting means late October and for Black Friday means the first week of November. The mechanics of upload timing, including why the day of the week matters less than creators think, are in the piece on the best time to post.
Google also publishes seasonal search data directly: its annual Holiday 100 gifting list is built from Search trends, and Trends shows where a category's interest sits week by week. Both beat guessing which products your niche will care about in November.
Packaging when every thumbnail turns red and green
There is a specific visual problem in Q4 that does not exist in any other quarter. Sometime in mid-November, a large fraction of the feed converges: red and green, gold foil, snow, bokeh lights, gift boxes, a bow in the corner. Every one of those creators is reasoning correctly — the seasonal cue tells the viewer this video is about now — and the aggregate result is a page of tiles that look like one another, on which nothing is distinguishable and the viewer defaults to the channel they already know.
The resolution is to treat the season as a signal, not a theme. A signal occupies a small, consistent part of the frame: a corner element, a colour accent in the text, one prop. The rest of the tile carries on doing what it does the other nine months — your colour, your type, your framing. Viewers recognise channels before they read titles, and December is the worst month to be unrecognisable to the people most likely to click you. The argument for holding a visual system in place is worked through in the piece on thumbnail consistency, and it applies double when everyone else is abandoning theirs.
Three concrete adjustments for the quarter:
- Differentiate on subject, not on palette. When ten tiles are the same colour, the one showing a clear object, a face with a readable expression, or a number wins the scan. Colour is the crowded channel in December; content is not.
- Assume a television. December is the month people watch on the family screen, and the living room is where the viewing records get set. Thumbnails that depend on a small detail or four words of fine text fail at ten feet. The specifics of how tiles render on TV are in the guide to optimising for TV screens.
- Cut text to three words. Seasonal titles are longer than usual ("last minute", "under $50", "2026"), so the thumbnail should carry less, not more. If the title is doing the date, the thumbnail should do the object or the verdict.
If you are producing a run of seasonal tiles in one sitting — a gift guide series, a twelve-video December schedule — the risk is that they end up either identical to each other or identical to everyone else. Generating variants and comparing them side by side at feed size, rather than designing one at a time at full size, is the cheapest way to catch both. That is what Thumblore is for: several packaged options for the same video in a few minutes, so the choice is made by comparison rather than by whichever one you made first.
Dated titles, evergreen files
A year in the title ("2026 Gift Guide") wins the click in season and kills the video in March. A guide with no year lasts, but loses to the dated one in the eight weeks that matter. The resolution most established channels arrive at is to date the title and treat the video as a yearly asset: publish dated in October, then next September change the title and thumbnail rather than filming it again from scratch. Title and thumbnail can both be edited on a live video without losing its watch history, and the ranking and CTR consequences of doing that — including when a swap costs you — are covered in changing a thumbnail after upload.
The same logic applies in reverse to seasonal styling in the frame. Snow and a bow date a thumbnail as firmly as a year does. If a video has a year-round life underneath its seasonal one, keep the snow out of the composition and put the seasonal cue somewhere you can edit out in February.
Testing inside a two-week shelf life
YouTube's Test & Compare runs up to three thumbnails against each other on a live video and calls the winner on share of watch time rather than clicks alone. It has one property that makes Q4 awkward: it needs impressions and it needs time, often up to a fortnight, to separate a real difference from day-of-week noise. A seasonal video may not have a fortnight. Three adjustments make testing work inside the quarter:
- Test on the evergreen videos, not the dated ones. Your best-performing library video gets more impressions in December than a new seasonal upload does, and the winner transfers to everything you package afterwards. The dated video benefits from the lesson; it does not have to be the laboratory.
- Run the test early in the season, not at the peak. A test on an October gift guide tells you which packaging direction works for the whole November run. The same test in the second week of December tells you something you can no longer use.
- Decide in advance what a result means. A win by a couple of points of watch-time share, called early on a small sample, is not a finding. The full protocol — sample size, when to stop, and which results to ignore — is in the Test & Compare guide and the broader thumbnail A/B testing system.
For anything you want to compare before publishing, the free thumbnail A/B comparison tool puts two candidates side by side at real feed sizes, which catches the December failure mode — two tiles that look different on your monitor and identical in a red-and-green feed — in about ten seconds.
The three lanes that do not reset on 1 January
Ad revenue is the lane with the most violent seasonality and the least control. The other three are negotiated, and the negotiating happens now.
Sponsorships. The Influencer Marketing Factory's 2026 brand deals analysis found the fourth quarter to be the most active period for creator collaborations across platforms, with YouTube taking the largest share of them. That demand is met with budgets that are committed weeks ahead: by the time November arrives, the slots are filled. If you want Q4 brand money, the conversations belong in September and early October, and the contract should specify a publication window rather than a vague quarter. The other half of that is disclosure — a paid integration has to be declared both to YouTube and to the viewer, and the rules and the penalty for getting it wrong are in the piece on branded content disclosure.
Affiliate commission. This is the one revenue line that rises with shopping intent rather than with advertiser budgets, which makes it the natural hedge against the January cliff — and it is the reason product-led channels often have a better Q1 than their RPM suggests. YouTube's Shopping affiliate programme lets eligible creators tag products directly in videos and Shorts and earn commission on what sells; figures reported from YouTube's own data put the lift from timestamped product tags at roughly 40% more product clicks than description links alone. Eligibility, the commission mechanics and the tagging workflow are covered in the guide to the Shopping affiliate programme.
Audience-funded income. Memberships, merchandise and anything your audience pays for directly are priced by your relationship with them, not by an ad auction, and they do not fall over on 1 January. December is a strong month to launch one, because gift-buying intent and goodwill are both high; the mechanics are in the piece on channel memberships.
Q5: the fortnight most channels sleep through
Retail marketers have a name for the period from roughly 26 December to mid-January: Q5. It is a behavioural quarter rather than a formal one, and Google's research into it is striking — 87% of shopping occasions in that window end in a purchase, higher than during the peak holiday period itself. People shopping after Christmas are spending gift money and returns credit on things they chose themselves. Intent is at its annual maximum.
Meanwhile the advertising competition has evaporated, which is precisely why your RPM is on the floor. For a creator, that combination is unusual and underused:
- New devices need content. Every phone, console, camera, television and piece of kit unwrapped on 25 December creates a person who wants setup guides, first-week tips, accessory recommendations and "I just got this, now what" videos. If your niche has hardware in it at all, this is the single most predictable demand spike of your year, and almost nobody publishes into it because they are on holiday.
- Resolution demand is real and short. Put fitness, personal finance, language learning or productivity into Google Trends and the annual crest sits in the first days of January every year. The videos have to exist before the 1st, not be filmed on it.
- The competition is asleep. Upload volume in most niches thins out over the holidays while viewing hits its annual peak, so a video published on 28 December competes with fewer new uploads for the same impressions.
- It is the right window for a swing. A format you have never tried, a longer video, a subject outside your lane — the cost of a flop is lowest when revenue per view is lowest anyway, and the audience is still large enough to give you a real read.
Practically: film the Q5 slate in the third week of December and schedule it. The work happens before the holiday, the publishing during it.
January, and how not to misread it
Three things happen in the same fortnight and creators routinely blame them on each other: revenue per thousand views falls, the channel's December uplift disappears, and the first payment of the year looks small. Take them in order.
The RPM fall is the auction, and it has nothing to do with your videos. The lost uplift is the seasonal audience going home — if a third of your December views came from gift-guide search demand, that traffic does not exist in January and its absence is not a decline in your channel. The best way to tell the difference is to open Analytics by traffic source and compare like with like: subscriber and browse traffic in January against the same in November. If those held and only search fell, the calendar did it. The full diagnostic for a drop that is not seasonal is in why did my YouTube views drop.
The payment timing catches people every year. Under the published AdSense timeline, a month's earnings are finalised and posted to the balance between the 7th and 12th of the following month, and payment is issued between the 21st and 26th if the balance clears the threshold. December's strong month therefore arrives in late January, and January's weak month arrives in late February — which is why the cash-flow dip lands a month after the revenue dip, usually in the week creators have already convinced themselves something is broken.
January is for everything you could not afford to do in December. Rebuild the packaging on your ten best library videos. Run the tests that need two weeks. Refresh last year's seasonal titles and thumbnails so they are ready for next October. Film the backlog that gives you slack in Q4 next time. The month is cheap in revenue and rich in attention, which is the right condition for work that pays later.
The one number to write down before you forget
Open Studio's revenue tab now and record your RPM for the first week of October, then again for the second week of December, then for the second week of January. Three numbers, your channel, your audience. Next year that beats every published CPM table in existence, and it takes four minutes.
A twelve-week plan
| Weeks | Priority | The thing that is easy to skip |
|---|---|---|
| Mid-September | Pick the two or three seasonal videos that will carry the quarter; open sponsor conversations | Check last year's Studio data for which videos actually earned, not which got views |
| Late September | Film the October and November slate; build the packaging system for the quarter | Decide now where the seasonal cue lives in the frame, so every tile is consistent |
| Early October | Publish the first buyer's-guide content into rising demand | Start a Test & Compare on an evergreen video to learn what wins before it matters |
| Late October | Publish dated gift guides; refresh last year's seasonal videos with new titles and thumbnails | Confirm sponsor publication dates in writing |
| Early November | Black Friday content live before the week itself | Product tags and timestamps on anything shoppable |
| Late November | Deals, reactions, fast turnaround | Do not start format experiments in the highest-paid fortnight of the year |
| Early December | Last-minute guides and your strongest evergreen swing | Check every thumbnail at TV size, not monitor size |
| Mid-December | Film and schedule the Q5 slate | Schedule it before you stop working, not after |
| Late December to mid-January | Publish into Q5: new devices, best-of, year-ahead, resolutions | Ignore the RPM number entirely for four weeks |
None of this requires more uploads than you already make. It requires the same uploads in a different order, with the expensive ones landing where the money and the demand overlap, and the experimental ones landing where a failure is cheap.
The deeper point is that the quarter rewards preparation more than effort. The auction pays the same premium to a video made in a hurry as to one made in October — but only if that video is in front of people, and being in front of people in December was decided six weeks earlier. Creators who have a good Q4 are rarely working harder in December. They are working in September and resting in the week the rest of the feed is scrambling.
Two things carry over into every month, seasonal or not: whether someone can tell at a glance that a video is yours, and whether the tile earns the click on a screen the size of a playing card. If you are building a run of seasonal packaging and want variants to compare rather than one guess, Thumblore generates them from your footage and your channel's look in a couple of minutes. If you would rather start with fundamentals than the calendar, the complete thumbnail playbook and the piece on what counts as a good CTR make everything above easier to judge.