Key takeaways
- Selling your own products and tagging other people's are two different features with two different gates. Almost every conflicting subscriber number online comes from writers confusing them.
- Your own store shelf unlocks at the Partner Programme's lower tier — 500 subscribers plus a watch-time or Shorts-views condition — not at 10,000, which is where the old merch shelf sat.
- No YouTube fee appears in any store partner's price list. The money leaves through your store subscription, your payment processor and your supplier's base cost, and that is where the margin is decided.
- One buyer per 2,000 viewers is roughly what it takes to match the ad revenue of the same video. Merch fails on margin and fulfilment far more often than it fails on traffic.
- Spring's shutdown this autumn is the lesson: order history does not migrate, store URLs stop resolving, and nobody hands you your customer list. Own the domain and the email addresses.
- A shelf tile is a thumbnail. It renders at a few hundred pixels wide next to your video, which means the same shrink rules that govern your packaging govern your product photography.
On 5 November 2026 the Creator Dashboard at Spring — the platform most YouTube creators knew as Teespring, and one of the first that the merch shelf ever supported — is scheduled to close for good. Its parent, Amaze Commerce, had already switched off stores and selling tools on 28 October, after a first round of notices in September pointed creators at a 15 September cut-off. Creators spent October exporting artwork and discovering that their order history was not coming with them. If part of your plan for this quarter was to sell merch on YouTube, the ground moved under the question before you got to ask it.
That is an unusually clean illustration of how creator merch actually works, which is not the way it is usually written up. The guides treat it as a feature to switch on: clear a threshold, connect a store, products appear under your videos, money arrives. The switch is real and takes about twenty minutes. Everything that decides whether it earns anything sits either upstream of it, in the margin on a physical object, or downstream of it, in the video that has to make somebody want the object in the first place.
This piece is the part the setup tutorials skip: which gate you are actually standing at, what a connected store commits you to, the arithmetic that tells you whether a drop is worth the week it costs, and why the product photo you upload obeys the same rules as the thumbnail sitting next to it.
Two features, one confused name
Search for merch requirements and you will be told you need 500 subscribers, 1,000, 10,000 or 15,000, often on pages published the same month. The numbers are not wrong so much as attached to the wrong things. YouTube Shopping is two separate capabilities wearing one brand name.
The first is promoting your own products from a store you control. You connect a store, YouTube pulls the catalogue, and your items become taggable in videos, Shorts and live streams, with a store shelf on your channel. This is the descendant of the merch shelf, and it is what people mean by selling merch on YouTube.
The second is the Shopping affiliate programme, where you tag other companies' products and earn a commission on sales that happen on their sites. That has its own eligibility rules, its own country list — YouTube said at Made on YouTube in September 2026 that it intends to reach 35 markets by the end of the year — and no inventory or fulfilment of any kind. The mechanics are different enough to deserve their own piece, and they have one: the Shopping affiliate programme at 500 subscribers.
The confusion matters because the two features fail for different reasons. Affiliate income fails when the chain from impression to purchase has a weak term in it. Your own merch fails when the unit economics do not survive shipping, or when you have an audience that watches you but does not identify with you.
Which gate you are actually standing at
For your own products, YouTube's own documentation leads with Partner Programme membership rather than a subscriber count, and that is the right way round. The expanded Partner Programme has two tiers, and merch sits in the lower one:
| Tier | Requirements | What it unlocks |
|---|---|---|
| Fan funding tier | 500 subscribers, 3 valid public uploads in 90 days, and either 3,000 valid public watch hours in 12 months or 3 million valid public Shorts views in 90 days | Memberships, Super Thanks, Shopping — including your own store |
| Advertising tier | 1,000 subscribers, and either 4,000 valid public watch hours in 12 months or 10 million valid public Shorts views in 90 days | Ad revenue share on top |
So the honest answer to "how many subscribers do I need to sell merch on YouTube" is 500, with a watch-time condition attached, and not the 10,000 that circulated for years because that is where the merch shelf used to sit. The watch-hour side of the advertising tier doubles on 1 February 2027, which is covered in the 2027 monetisation requirements; the fan-funding thresholds are unchanged by that, so the merch door is the one that is getting relatively easier.
Three more conditions sit alongside it. Your channel has to be in a supported country, and the country list for selling your own products is not the same as the affiliate list — it changes several times a year, which is why no article is a reliable source for it. Your channel cannot be set as made for kids, and individual videos marked made for kids will not carry product tags either, for the same reason those videos lose most commercial features; the setting is explained in the made-for-kids guide. And the usual good-standing requirements apply.
Do not take any of that from an article, including this one. Open YouTube Studio, go to Earn, and look for a Shopping tab. If it is there, you are eligible today, in your country, on your channel. If it is not, nothing you read about thresholds will change that, and the fastest route is usually the 500-subscriber tier rather than an appeal.
What a connected store actually is
You do not sell on YouTube. You sell on your own store, and YouTube displays a window into it. That distinction explains most of the friction creators hit.
Your catalogue has to live on a supported platform — Shopify has the deepest integration and is the most commonly connected, with Fourthwall, Spreadshop and a handful of regional platforms also supported depending on where you are. The supported set differs by country and gains and loses names, so check Studio rather than a comparison post.
Once connected, your products go through two reviews, not one. They are checked against Google Merchant Center's listing policies and against YouTube's own policies, and an item that fails either set is simply hidden from Studio rather than flagged with an explanation in front of you. Review runs in business days, not minutes. When something is disapproved, the fix happens in Merchant Center on the store side — the retailer or store administrator resubmits it — which means a creator who connected a store built by somebody else has to go back through that person to unblock a listing. Google's Shopping policies and requirements are the authority on what trips it, and the useful habit is to read the specific policy reason rather than resubmitting the same listing twice.
On the display side, your store shelf appears at the top of your channel homepage by default. That position is yours to change or switch off in Studio, and it is worth a thought rather than a shrug: a shelf above your video rows is a shop that greets every new viewer before your work does.
The twenty-minute setup is not the project
Connecting a store, choosing which items appear and tagging your first video genuinely takes under half an hour. Designing something people want, getting a sample made, photographing it, pricing it so the margin survives shipping, and building the video that sells it is a week. Budget for the week, not the half hour, and the half hour will feel like what it is — the last step.
What Spring's shutdown should teach you about where your store lives
Spring was not a marginal player. It was Teespring, one of the first integrations YouTube supported, and for years it was the default answer for a creator who wanted shirts without inventory. Amaze Commerce's own Spring update page is the source for the timeline: stores and selling tools switched off on 28 October 2026, with the Creator Dashboard scheduled to close on 5 November 2026, and creators told to download artwork, images and listing details before the selling tools went dark. Invited creators were offered a free migration to Amaze Commerce that recreates up to five best-selling products and moves eligible artwork and brand assets.
Read what that offer does not include, because the pattern repeats every time a creator platform winds down:
| Asset | What happens when a store platform closes |
|---|---|
| Store URL | Stops resolving. Every description link, pinned comment and end screen pointing at it becomes a dead end, including on your back catalogue. |
| Order history | Does not migrate. Amaze states this plainly for Spring. Export it before you need it for returns, tax or support. |
| Product listings | Do not transfer automatically. Even a generous migration rebuilds a handful by hand. |
| Artwork and source files | Yours only if you still have the originals. Download the print-ready files, not the previews. |
| Customer relationship | Whoever holds the email addresses holds it. On most POD platforms that is not you by default. |
The practical conclusions are small and boring. Point your audience at a domain you own, redirecting to whichever platform you currently use, so a migration costs a DNS change rather than every link you have ever published. Keep print-ready artwork in your own storage, named so you can find it in two years. Export orders and any customer list you are permitted to hold, on a schedule. And treat a merch platform the way you treat a video host you do not control: fine to use, not fine to be the only copy.
The arithmetic nobody runs first
Merch gets discussed as a revenue stream and priced as a hobby. The number that matters is margin per unit, and the number that decides whether a drop was worth doing is units sold multiplied by that margin, against the hours the drop cost you.
Start from the comparison every creator actually wants. Take a video that will finish on 50,000 views, on a channel earning a $5 RPM — so roughly $250 in advertising from that video. Suppose your profit on a shirt, after the supplier's base cost and payment processing, is $10. Matching the ad revenue takes 25 shirts. That is one buyer per 2,000 viewers, or 0.05% of the audience.
Stated that way merch sounds easy, and the arithmetic is honest: one in two thousand is a far lower bar than most creators assume. Three things make it harder, and all three sit on the cost side rather than the demand side.
The margin is thinner than the price suggests. Print-on-demand economics are roughly: supplier base cost, plus processing at something like 2.9% plus a fixed fee per transaction — Fourthwall publishes 2.9% + $0.30 for US cards and 3.9% + $0.30 for international ones, and most processors are in that neighbourhood — plus whatever your platform charges. On catalogue items Fourthwall deducts the base cost and leaves you the markup rather than taking a percentage. YouTube's own cut does not appear anywhere in these fee schedules: no store partner's price list carries a YouTube percentage on an own-product sale, which is consistent with the shelf being a window into your store rather than a checkout. A $28 shirt against a $14 landed cost does not leave $14; it leaves something closer to $12 before a single return, and a return usually costs you the whole unit.
Shipping is where drops die. A $9 delivery charge on a $28 shirt kills conversion more reliably than any design decision, and international shipping on physical goods can exceed the item's margin outright. If a meaningful share of your audience is outside your supplier's cheapest shipping zones — and on YouTube it almost always is — check the delivered price in three countries before you commit to a product, not after.
Units do not scale with views the way ad revenue does. Advertising accrues quietly on the back catalogue for years. A drop is front-loaded: most of its sales land in the first week, driven by the videos that announce it, and then it goes quiet. One good drop a quarter is a different business from a monthly RPM, and the comparison above flatters merch by ignoring that the 25 shirts have to be sold again next time.
Why nobody can tell you your conversion rate
Every merch guide quotes a conversion figure. The ranges you will see — commonly 0.5% to 3% of an audience per drop — come from print-on-demand vendors' own blog posts, not from audited data, and the sources disagree with each other while citing nothing. There is no platform-published figure for what share of a channel's viewers buy a shirt, and there is a good reason for that: the variance between channels is larger than the number itself.
So measure instead of forecasting. One drop, one product, one announcement video, and you have your own rate — expressed as units per thousand views on the announcing video, which is the only denominator you can act on. Studio's analytics will show you how many viewers interacted with the shelf; your store will show how many of those bought. Two drops give you a baseline that is worth more than every benchmark table on the internet, because it is measured on your audience rather than somebody's client list.
Until you have that, plan with the break-even question rather than a forecast: at my margin, how many units pay for the week this costs me? If the answer is a number you would be surprised to hit, the product or the price is wrong, and promotion fixes neither.
Density beats size
The pattern that holds across channels that sell merch well is not subscriber count. It is whether viewers watch you as a membership or consult you as a reference.
A viewer buys a shirt to say something about themselves. That is why identity-driven niches — gaming communities, fitness and running, music, motorsport, the long-running podcast with its own vocabulary — convert far better than utility niches at the same size. Nobody signals membership of a spreadsheet tutorial. The inside joke from episode 40 of a series somebody has watched all of is a wearable statement; a tidy logo for a channel they watched twice is a garment with a stranger's branding on it.
Two consequences follow. First, if your channel is reference-shaped — tutorials, reviews, how-to, explainers — your audience-funded money is more likely to come from memberships, Super Thanks or digital products than from apparel, and that is not a failure of your merch design. Second, if your channel is identity-shaped, the design brief is not "put the logo on it". It is "make something a fan would wear in front of people who are not fans", which is a much harder and much more valuable brief. Gaming channels in particular have a head start here and a trap: the iconography that would actually sell often belongs to a publisher, and fan-content permissions that allow a thumbnail do not allow merchandise. The Fortnite thumbnail piece walks through where that line sits for one publisher, and the same reasoning applies to every other.
A shelf tile is a thumbnail
Here is the part no merch guide covers, and it is the part this site is for. When your product appears under a video or on your channel, it renders as a small tile beside your thumbnail — a few hundred pixels wide on a desktop, smaller on a phone, in a row a viewer scans rather than studies. Which means your product photography is governed by the same constraints as your packaging, and most creators upload images that were designed for a product page viewed at full width.
The failures are the familiar ones. A white shirt photographed on a white background has no edge, so the tile reads as an empty rectangle. A full-length mockup puts the design at perhaps a fifth of the tile's height, which is illegible at tile size, so the viewer sees a person rather than a product. A design with small type on it — a quote, a date, a handle in a thin weight — disappears entirely, the same way thin type disappears from a thumbnail. The rules worked out in the thumbnail fonts piece about cap height as a fraction of the frame transfer directly: if the words on the garment have to be read, they have to be big relative to the image, not relative to the garment.
What works is the thumbnail answer applied to products. Crop in so the design occupies most of the frame rather than shooting the whole garment. Shoot against a tone that separates from the product — the same figure-ground reasoning as thumbnail colour. Keep one subject per tile, because a tile is scanned in the same fraction of a second as a thumbnail and composition under crowding behaves the same way. And check the result at real size before you publish it: the thumbnail preview tool renders feed sizes, and if your product tile is unreadable next to your own thumbnail, viewers never see the product you spent a week designing.
One more specific: shelf tiles are roughly square, your product photos are probably not, and automatic cropping will take the decision away from you. Shoot or crop deliberately for a square, keeping the design centred, and use the resizer rather than letting a platform guess.
The drop is a video, not a shelf
The shelf converts intent it did not create. Something has to create the intent, and on YouTube that something is a video.
Channels that sell merch successfully do roughly this. The product appears in content before it is for sale — worn, used, referenced — so that by launch day the audience has seen it three times. The announcement is a video with a reason to exist beyond the announcement, because a pure advert underperforms and sits badly in the feed. The callout is early and brief, in the first two minutes rather than buried at nine, and repeated once at the end. The shelf is tagged on that video and on the two or three back-catalogue videos that still pull traffic — which is where most of the sales actually come from over the following month, not from the launch video. An end screen and a community post carry it after the first week.
Timing is worth one sentence of planning. Gift-buying intent in November and December is the strongest demand signal a creator gets all year, and a drop timed for it has a different ceiling from the same drop in February; the rest of that seasonal argument is in the Q4 strategy piece. Set your own deadline backwards from it: samples take weeks, and a product that arrives on 20 December arrives after the money has been spent.
If you are pairing the launch with a giveaway, the rules on contests are stricter than most creators realise; the giveaway guide covers them.
The compliance work that arrives with physical goods
Selling an object is a different legal category from selling attention, and two obligations surprise creators.
The first is product safety law if you sell into the European Union. The EU's General Product Safety Regulation — Regulation (EU) 2023/988, applicable since 13 December 2024 — prohibits placing a product on the EU market unless an economic operator established in the Union is responsible for it. Under Article 16 that responsible person is an EU-established manufacturer, or an importer, or an authorised representative holding a written mandate, or an EU-established fulfilment service provider where none of the others exists; their identity and contact details have to appear on the product, its packaging, the parcel or an accompanying document. A Great Britain address does not satisfy it. In practice, print-on-demand suppliers with EU production will often supply the product-level safety information while leaving the responsible-person details to you, which is a sentence worth reading on your supplier's help pages before you take EU orders rather than after.
The second is more mundane: the designs themselves. A logo, a character, a club crest or a game's iconography on a shirt is commercial use of somebody else's mark, and the fan-content allowances that cover a thumbnail or a video almost never extend to merchandise. If a design element did not come from you or from a licence you can point at, it does not belong on a product, however well it would sell.
When not to sell merch on YouTube
Three situations where the answer is no, or not yet.
If you have not published anything an audience would quote back to you, you do not have merch yet — you have a logo. The design that sells is downstream of content that created a shared reference, and there is no shortcut to the reference.
If your audience is overwhelmingly in regions where delivered cost is punitive, the product will convert and then disappoint, and a bad first order costs you more than the margin on it. Digital products, memberships or affiliate tags fit better.
And if your packaging is the thing holding the channel back, merch is a distraction with a production schedule attached. A week spent on click-through on your next five uploads compounds; a week spent on a shirt pays once. The ordering question is answered by your own numbers — if your click-through is below where comparable videos on your channel land, as discussed in what counts as a good CTR, fix that first.
The bottom line
Merch on YouTube is retail with a distribution advantage, and it is priced accordingly: no YouTube percentage, your supplier and processor taking theirs, and what is left is a real margin on a real object you are now responsible for. The gate is lower than almost everybody thinks — the 500-subscriber tier, not 10,000 — and the setup is the easiest part of the whole exercise. The hard parts are a design people would wear in public, a margin that survives shipping, and an announcement video that earns the attention it spends.
Spring's closing this week is the argument for doing the unglamorous version: own your domain, keep your artwork, export your orders, and treat any platform as replaceable. Creator infrastructure changes faster than creator audiences do, and the assets worth protecting are the ones a shutdown email cannot touch.
The other thing worth protecting is the traffic all of this multiplies. Every term in the merch arithmetic starts with somebody opening a video, which is the term a thumbnail decides — and the one place where a week of work keeps paying after the drop is over. If that is the part you want to improve first, Thumblore turns a described thumbnail into options you can test in an afternoon, and the gaming thumbnail generator is tuned for exactly the identity-driven niches where merch tends to work. Sell the shirt by earning the click.