Creator economy · Link mechanics
Does a product URL change when a size sells out?
Four fifths of the time a size is its own page, which decides whether ‘link the item or the set’ is a choice at all.
Some links here are affiliate links — if you buy, the retailer may pay us a commission at no extra cost to you. We have earned $0 from them so far. Here’s how it works.
Part of: For creators
Yes, usually. Of 136 multi-variant product groups in a 2,320-listing pull on 3 September 2026, 111 — 82% — gave every single variant its own URL. Only 22 put the whole group on one address. When a size has its own URL, that URL is what dies when the size does.
So the answer to the practical question is: link the product, not the size, unless the size is the entire point of the recommendation. At most retailers those are two different addresses.
What the feeds actually do
Product feeds carry a field called itemGroupId, which is how a retailer says “these rows are the same product in different sizes or colours”. Reading it tells you exactly what you want to know, and it is rarely read.
| Measurement | Count | Share |
|---|---|---|
| Listings in the pull | 2,320 | — |
| Carrying a variant group id | 937 | 40.4% |
| Carrying a size | 1,613 | 69.5% |
| Carrying a colour | 1,622 | 69.9% |
| Distinct product groups | 592 | — |
| Groups with more than one variant | 136 | — |
| … where every variant has its own URL | 111 | 82% of those groups |
| … where all variants share one URL | 22 | 16% |
| Largest single product group | 86 variants | — |
Four fifths of the time, a size is a separate page. That is the finding, and everything else on this page follows from it.
It is a per-retailer decision, and it is all or nothing
The most useful thing in this data is not the average. It is that there is no middle ground: retailers either publish variant grouping on everything or on nothing.
| Retailer | Listings | Carrying a variant group id | Carrying a size |
|---|---|---|---|
| Wayfair North America | 700 | 100% | 98% |
| OnBuy.com | 487 | 0% | 5% |
| Poshmark | 365 | 0% | 100% |
| Wayfair UK | 290 | 0% | 100% |
| Perigold | 171 | 100% | 100% |
| OnBuy EU | 169 | 0% | 0% |
| SHEIN | 68 | 0% | 99% |
| Qathu - K-Beauty Store | 51 | 100% | 0% |
Every cell in that middle column is 0% or 100%. Nothing sits between.
Look at the two Wayfair rows in particular. The same brand, in two markets, behaves in opposite ways — the North American feed groups every listing, the UK feed groups none. That is not a policy difference about products; it is a difference between two feed exports. It is also a warning against reasoning about “what this retailer does” from one market’s catalogue.
Keep reading
What breaks, and how it looks when it does
Three distinct failures come out of this, and only one of them looks like a failure.
The size sells out and the URL 404s
The clean case. Your link is dead, a checker catches it, you fix it. Annoying and visible.
The size sells out and the URL redirects to the product
Better for the shopper, worse for you: the link returns 200, so nothing flags it, and every reader now lands on a page where the thing you actually recommended is unavailable or shown at a different price. Our own 80-URL trace found redirects that dropped query parameters while returning 200, which is the same shape of problem — a healthy status code hiding a changed outcome.
The size sells out and the page stays up, greyed
The worst one. Live URL, live page, unbuyable product, and a reader who concludes you did not check. Nothing automated will ever tell you.
Two of those three pass a link check. That is why “my links all return 200” is a weaker statement than it sounds.
The rule this produces
- Link the product page, not the variant, unless the specific size or colour is the reason you are recommending it.
- If the variant is the point, say so in the caption. “The 8 is the one that fits true” survives the link dying; a bare link does not.
- Treat variant-heavy categories as higher maintenance. Clothing and footwear churn URLs in a way a wall clock does not, because there are more addresses per product and each one can die alone.
- Re-check clothing links on a shorter cycle than homeware or electronics. Not because retailers are worse, but because there is simply more to go wrong per product.
- Check the page, not the status code. Two of the three failure modes above return 200.
What we could not observe
- We did not watch a size sell out. That needs the same URLs re-checked over weeks, and this is a single-day snapshot. What we measured is the addressing: whether a size has its own URL to lose. We have taken the baseline and will publish the decay when a second measurement exists.
- We cannot see the retailer’s behaviour on sell-out — 404, redirect or greyed page — from feed data alone.
- Six categories, one morning. Variant behaviour in clothing specifically may differ from the homeware-heavy mix here.
- A group id is the retailer’s claim, not a fact. A retailer that does not publish one may still group products internally.
Frequently asked
Does a product URL change when a size sells out?
At most retailers the size has its own URL to begin with, so it is not that the URL changes — it is that that URL stops working while the product’s main page carries on. In a 2,320-listing sample, 111 of 136 multi-variant groups gave every variant a separate address.
Should I link a size or the product?
The product, unless the size is the recommendation. A product page survives a single variant going out of stock; a variant URL does not.
How can I tell whether a link is a variant URL?
Strip the query string and reload. If a different colour or size appears, the parameter was load-bearing and you were on a variant URL.
Do all retailers handle variants the same way?
No, and there is no middle ground. In our sample every retailer published variant grouping on either 0% or 100% of its listings, and the same brand behaved oppositely in two markets.
Which categories break most often?
The ones with the most variants per product — clothing and footwear — simply because there are more separate addresses to lose.
Method and sample
- Sample: 2,320 unique live listings across six seed categories, pulled 3 September 2026 through the CJ shopping-product API with the variant fields selected.
- A note on those fields. Two earlier runs of ours recorded that variant and availability fields could not be selected at all. That was wrong, and this study exists because we re-tested it: the fields are selectable, and
targetCountryonly fails when used as a filter. - A “group” is all listings from one retailer sharing an
itemGroupId. Groups of one are excluded from the URL comparison. - Two of eight seeds failed on a DNS error and are absent from the sample.
- Baseline logged: 3 September 2026. The decay half of this question needs a second date and does not exist yet.
Keep reading
Last verified 3 September 2026 against 2,320 live listings read for variant grouping, size and destination URL
How this guide was made. We research and draft these guides with AI, then a person checks every price, link and factual claim against the source before it publishes. We work this way because it lets us re-verify prices across hundreds of guides in a day, which is what keeps the numbers here current; it does not decide what we recommend. Anything we could not verify is labelled as unverified rather than filled in.