Merchentia.

Sold out, still shoppable: how stock-aware looks maintain themselves

A practical guide for merchandising & ecommerce ops teams · updated September 2026

A curated outfit is only as good as its least-available item. One sold-out bag turns a styled look into a dead end — and on a hand-built look, the only thing standing between that and a lost sale is a person who notices. A complete-the-look widget that checks stock itself removes that person from the loop entirely.

The maintenance cost nobody budgets for

Curated looks are easy to launch and expensive to keep. Every look you hand-build is a small standing commitment: when something in it sells out, someone has to notice, pick a replacement, and edit the look. Ten looks is a habit. Five hundred is a recurring task on someone’s board. A full catalogue of them is a job.

What makes it worse is that the failure is silent. Nobody files a ticket saying “the look on product 4821 is recommending a bag we haven’t been able to ship since Tuesday.” The shopper just hits a sold-out card, loses the thread, and leaves — and the only trace is a conversion number that’s slightly worse than it should be.

What “stock-aware” actually means

The fix isn’t a person checking looks more often. It’s a look that was never a fixed list of products to begin with.

Each slot in the outfit — the shoes, the bag, the jacket — holds a ranked list of candidate matches from your own catalogue, not one hard-coded product. Which candidate a shopper actually sees is decided at the moment the widget loads, not when the look was created. That single design choice is what turns stock handling from a chore into a non-event.

What happens the moment something sells out

  1. The sold-out item drops out of the running for its slot — on the very next page view, with nothing regenerated.
  2. The next-best match takes its place. Because the candidates are ranked by how well they fit that slot, the replacement is the closest remaining alternative, not arbitrary filler.
  3. If nothing in that slot is buyable, the slot disappears. The shopper sees a shorter outfit rather than a card they can’t act on — the rest of the look keeps working.
  4. When it’s restocked, it becomes eligible again on its own. There is no rebuild and no re-analysis to trigger.

The same check covers a case merchants often miss: a product you unpublish or archive has no product page at all, so recommending it would send shoppers to a 404. Buyability here means active, in stock, and still published to the Online Store channel — all three, re-checked together.

Why this is an operations story, not a feature list

The interesting part isn’t that sold-out items get swapped. It’s which work stops existing:

That last point is the one that changes staffing maths. Hand-curated merchandising scales linearly with catalogue size — more looks, more upkeep, eventually more people. Serve-time buyability doesn’t scale at all, because there is no per-look maintenance to scale.

The honest limits

Two things worth knowing before you assume it covers everything:

It works at product level, not variant level. A jacket with five sizes where one is sold out is still a buyable jacket, so it stays in the look — which is usually what you want, since the shopper can still purchase it. A product only drops out when the whole product is unavailable.

Stock has a backstop, not a live feed. Unpublishing or archiving a product takes effect right away. A pure inventory movement — the last unit selling, a manual restock — is picked up by a periodic catalogue re-sync that runs every few hours, because the app deliberately asks for read-only product access and no inventory permissions. If you’d rather not wait after a big restock, a manual re-sync is one click. It’s a narrower permission footprint in exchange for a short delay on one edge case.

Do it with Shop The Look

Merch - Shop The Look works exactly this way. AI detects the outfit in the product photos you already have and matches each item to your own catalogue, then re-checks buyability on every single widget load — serving the closest available match, dropping a slot when nothing in it can be bought, and letting restocked items return on their own. You keep full control to review, edit, reorder or swap any look before it goes live, and a “you may also like” related-products widget is included on every plan. It installs as a no-code theme app block, stores no shopper personal data, and the AI analysis runs ahead of time so the storefront only ever loads pre-computed results. Live on the Shopify App Store, with plans from a free tier up to Enterprise.

Install on the Shopify App Store →

About Shop The Look →

Frequently asked questions

Will the widget ever show a sold-out product?

Not in the normal path. Buyability is re-checked every time the widget is served, not baked in when the look is built, so a sold-out item is replaced by the next-best available match rather than shown.

What if nothing in a slot is available?

That slot is dropped from the outfit entirely. The shopper sees a shorter look rather than a card they can’t buy, and the rest of the outfit still works.

Do I have to rebuild looks when inventory changes?

No. A look is analysed once; which item actually shows is decided at serve time. Nothing is regenerated when stock moves, and an item becomes eligible again on its own once it’s back in stock.

Is stock checked per size or per product?

Per product. A product with some sizes sold out is still buyable, so it stays in the look; it only drops out when the whole product is unavailable, archived, or unpublished from the Online Store channel.

How quickly does a sell-out take effect?

Unpublishing or archiving a product takes effect right away. A pure stock movement, such as the last unit selling, is picked up by a periodic catalogue re-sync every few hours, and a manual re-sync is available if you don’t want to wait.

More from Merchentia

See all apps →

Questions?

Email ask@merchentia.com — we read everything.