Shorts pay 3–14% of what long-form earns per 1,000 views. That's not the interesting part — the interesting part is why chasing that number directly is, in the data's own words, the wrong thing to optimize for.
Source data published June 2026 · RPM figures shift — re-verify before publishing
Shorts generate 3–14% of what long-form earns per 1,000 views.
For most channels, that's under 2% of total revenue. The real value is somewhere else entirely.
Most "Shorts RPM by niche" content you'll find is one of three things: a creator's own Studio screenshot, a third-party tool extrapolating from public view counts, or an advertiser-side estimate describing what brands pay to run campaigns — not what a creator actually receives. All three skew high, and none of them isolate what Shorts specifically do to a channel's revenue.
| What this study did differently | Detail |
|---|---|
| Real API data | 274 channels, 3,044 channel-month data points, accessed directly through the YouTube Analytics API — not estimated. |
| Full-year medians | No single viral month or Q4 spike distorts the figures. Published Studio screenshots tend to come from December, when CPM runs 40–60% above January. |
| Within-channel design | Comparing Shorts-posting channels to non-Shorts channels is contaminated — those are different audiences and setups. This study instead isolated 149 "irregular" channels that posted Shorts in some months and not others, comparing each channel to itself. |
Comparing a channel that never posts Shorts to one that does mixes in every other difference between those two channels — audience, content quality, monetization setup.
Comparing the same channel to itself, in months it posted Shorts versus months it didn't, removes that contamination. Whatever difference shows up is much more likely to actually be about Shorts.
Source: AIR Media-Tech, "YouTube Shorts RPM niche-by-niche: real monetization data from 274 channels," June 2026. Niches covered: News & Politics, Crafting and Handmade, Entertainment, Business & Finance, Gaming, Kids & Teens, Music, Education & Science, Lifestyle, Food & Cooking, Transport, Gadgets & Stuff, Health and Sport. Size tiers: small (10K–100K monthly views), medium (100K–10M), large (10M–50M).
The source measured a standalone Shorts RPM for some niches, and for others it only measured how Shorts affect long-form RPM — a different, related question. We're keeping those separate below rather than blurring them into one table, because presenting an inferred number as measured would be exactly the kind of overstatement this page exists to correct.
| Niche | Shorts RPM | Long-form RPM | Shorts as % of long-form |
|---|---|---|---|
| Music (small, owns rights) | $1.48 | — | 60% |
| Music (medium, owns rights) | $0.97 | — | 41% |
| Lifestyle | $0.16–$0.20 | $3.87–$4.88 | ~4% |
| Crafting & Handmade | $0.11 | $3.53–$3.68 | 2.9% — lowest ratio in the dataset |
| Kids & Teens (large tier) | $0.02–$0.05 | $0.49 | — |
| Niche | Long-form RPM (measured) | Shorts RPM (inferred by the source) |
|---|---|---|
| Education & Science | $14.97 (medium), $18.23 (small) — highest in the dataset | $0.54–$1.44, calculated by applying the general 3–8% ratio to Education's long-form RPM. The source states this explicitly as an estimate, not a direct measurement. |
Whatever Education's real Shorts RPM turns out to be, an $18 long-form RPM makes every view redirected to Shorts genuinely expensive in opportunity-cost terms. The source's own finding reinforces this: Education channels that went Shorts-heavy showed the sharpest subscriber underperformance of any niche in the dataset — a 3.3× subscriber gap at the largest channel tier.
Gaming and Entertainment are covered in this study, but only through their effect on long-form RPM when a channel adds Shorts — not through a separate Shorts RPM figure. That's real, useful data (see Part 5), just a different measurement than the table above.
Business & Finance, News & Politics, Food & Cooking, and Transport were flagged in AIR's companion long-form study as having fewer than 10 channels each — read any figures for these as directional at best. Health & Sport wasn't broken out with specific figures in what's publicly available from this report. If your niche isn't in the tables above, the honest answer is that solid public data doesn't yet exist for it — the general $0.01–$0.10 range from our Shorts monetization guide is the best fallback.
The practical question behind all of this: how many Shorts views does it take to earn what 1,000 long-form views would earn?
| Niche | Shorts views to match 1,000 long-form views |
|---|---|
| Music (owns rights) | ~1,100 — dramatically better than every other niche |
| Kids & Teens (large tier) | 2,156 — low bar, but only because long-form RPM is also very low |
| Education & Science | 12,500–33,000 (using the inferred RPM range) |
| General range, most niches | 11,000–34,000 |
| Lifestyle (medium) | ~26,000 — one of the worst exchange rates measured |
| Crafting & Handmade | ~34,000 — widest gap in the dataset |
Direct Shorts revenue accounts for under 2% of total revenue on most channels in this dataset — while often consuming a disproportionate share of production effort relative to that return.
That's not an argument against making Shorts. It's an argument against making them for direct revenue. Part 6 covers what they're actually good for.
Small Music channels that own their rights hit $1.48 Shorts RPM — 60% of their own long-form RPM. Medium channels hit $0.97, or 41%. No other niche comes remotely close on either number.
When a Short uses copyrighted music, the rights holder receives a share of the revenue through YouTube's Content ID system. A Music channel that owns its rights — or has an arrangement with a distributor — collects this on top of standard ad revenue. A Gaming channel, a Cooking channel, or an Education channel has no equivalent mechanism available to it.
The practical takeaway, direct from the source: if you're in Music and own your rights, Shorts' viability as a direct revenue stream is real in a way it isn't anywhere else in this dataset.
One caveat from the source's within-channel data: Music channels show higher revenue in months they post Shorts ($606 vs $175), but this is confounded — Music channels likely post Shorts during their most active, highest-revenue periods regardless of the Shorts themselves. The relationship isn't necessarily causal.
| Metric | Value |
|---|---|
| Cross-sectional revenue gap | 5.3× (no-Shorts channels at $2,979/mo vs. irregular-Shorts channels at $559/mo) |
| Within-channel long-form RPM, no Shorts | $4.99 |
| Within-channel long-form RPM, 6–20 Shorts/month | $2.59 |
| Within-channel long-form RPM, 21+ Shorts/month | $1.15 |
The source is explicit that this doesn't prove Shorts cause the RPM decline — but calls the pattern "consistent enough to flag as a real strategic risk." An Entertainment channel with a high long-form RPM baseline has real money at stake if it scales Shorts volume without watching its own Studio data.
| Metric | Value |
|---|---|
| Long-form RPM, no Shorts | $3.42 |
| Long-form RPM, with Shorts | $3.29 |
| Difference | 4% — within measurement noise |
The dose-response curve is odd, though: RPM drops to $1.25 at 6–20 Shorts/month before partially recovering to $2.11 at 21+ — likely because heavy-Shorts Gaming channels in the sample tend to be clip-focused channels with a different monetization profile, not general Gaming channels that added Shorts to an existing strategy.
| Metric | Value |
|---|---|
| Long-form monetized playback rate | 12.1% — only 1 in 8 views carries an ad, due to COPPA ad-serving limits |
| Shorts monetized playback rate | 45.7% |
| Large-tier Shorts revenue share at 21+/month | 41% — highest in the entire dataset |
Because COPPA suppresses long-form monetization so heavily, Shorts are proportionally more monetizable in Kids content than their raw RPM suggests. For large Kids channels that have already maximized long-form output, Shorts can become a genuinely meaningful part of total revenue — not just a discovery tool.
We don't use Shorts for direct revenue. Shorts expand who finds the channel. Long-form is where that audience generates revenue. — AIR Media-Tech
Views and recommendations had cooled through summer 2025. The creator was skeptical Shorts were worth the effort given known-low RPM. Posted the first Short anyway.
Revenue grew 1.5× faster than views — because the Shorts themselves generated almost nothing directly. What they did was put the channel back in front of new viewers, some of whom stayed and watched the existing long-form catalog.
Localization had already driven 400M+ views; that lever was exhausted. The team pushed Shorts from 6.2% to 15.7% of total traffic alongside packaging and metadata work.
Long-form uploads actually dropped 35% in the same period. Every point of that 153% revenue gain traces back to long-form sessions — the Shorts opened the door, the existing catalog did the earning.
Long-form output was deliberately cut nearly in half (21 videos to 11) while Shorts more than doubled (10 to 23), reformatted from existing long-form highlights.
Revenue grew 6× faster than views. A modest-looking view increase represented millions of additional views at that channel's scale — better-matched audience, more monetized watch time per view.
None of these gains came from Shorts revenue itself. In every case, Shorts expanded the audience finding the channel, and the existing (or lightly refreshed) long-form catalog converted that expanded reach into revenue at full long-form RPM.
| Question | What the data says |
|---|---|
| Is 1–5 Shorts/month safe? | Yes, in most niches — minimal RPM impact visible in the dose-response data. |
| What's the sweet spot ratio? | 0.28–0.40, roughly one Short for every two to three long-form uploads — shows the strongest combined subscriber and revenue performance across a wider 18,000-channel sample. |
| What does heavy posting (21+/month) cost? | Decreases revenue in most niches — Entertainment and Education show the clearest damage. |
| Are there exceptions to "heavy is bad"? | Yes — large Kids channels and clip-focused Gaming channels show different patterns at high volume, tied to their specific monetization structures. |
Across the wider sample, channels that shifted the majority of their output to short-form showed decline on both subscribers and revenue — not just a lower ceiling, an actual regression from where they started.
Same discovery-first approach this data supports — see the full workflow.
Read the VidIQ guide →| Mistake | Consequence |
|---|---|
| Optimizing for direct Shorts RPM | The source's own conclusion: this is the wrong thing to optimize for in almost every niche. |
| Trusting a Shorts RPM figure above ~$0.25 as typical | Real measured figures top out around $1.48, and only for Music with owned rights. |
| Assuming Education's inferred Shorts RPM is measured | It's the source's own calculated estimate, not a direct measurement — treat accordingly. |
| Avoiding Shorts because RPM is low | Misses the actual mechanism — Shorts expand discovery, long-form converts it. |
| Posting 21+ Shorts/month by default | Correlates with revenue decline in most niches, especially Entertainment and Education. |
| Comparing your channel's Shorts RPM to Music's | Music has a mechanism (Content ID) unavailable to almost every other niche. |
| Panicking over Entertainment/Education RPM correlation | The source explicitly states causality is unproven — a real signal to watch, not a certainty. |
| Ignoring your own niche's monetized playback rate | Kids content proves this matters — 12.1% long-form vs. 45.7% Shorts, driven by COPPA specifically. |
This data describes AIR Media-Tech's specific 274-channel sample — real and rigorously measured, but not a census of all YouTube. Education & Science's Shorts RPM is explicitly an inference by the source, not a direct measurement, and is labeled as such throughout this page. Four niches (Transport, News & Politics, Food & Cooking, Business & Finance) are flagged by the source's companion study as having fewer than 10 channels; treat any figures for them as directional. Health & Sport is named in the study's niche list but wasn't broken out with specific figures in what's publicly available. Correlational findings (Entertainment and Education RPM decline with Shorts volume) are explicitly presented by the source as unproven causation, not certainty. RPM figures shift over time — re-verify before publishing anything that states a number as current.