Consignment in luxury used to mean a favor. A stylist pulled four pieces for a top client, sent them to her home with a handwritten note, and kept track of what went out on a card in a drawer. If she kept two, someone rang them up when she next came in. If she kept none, someone remembered to put the pieces back on the floor. Usually.
Today that favor is a service clients expect, and it sits next to a second one: reserving a piece before they travel to the boutique, confident it will be there when they arrive. Together, reserve orders and consignment on approval are how luxury retail bridges the gap between discovering a piece online and buying it in person.
The pressure to offer both is structural. The National Retail Federation and Happy Returns expect US shoppers to return $849.9 billion of merchandise in 2025, 15.8% of sales, and 19.3% of everything bought online, according to the 2025 Retail Returns Landscape. For a luxury house, asking a client to buy five figures of garments online, wait for delivery, send most of it back and wait again for a refund is not a service model. It is a reason to shop elsewhere.
The catch: most retailers offering holds and approvals are doing it with technology that was never designed for them. A generic point-of-sale system assumes a unit is in one of two states: on the shelf and for sale, or sold. A held piece is neither. A piece out on approval is neither. And when that assumption breaks, the stock figure every channel relies on breaks with it.
The retailers scaling this well are not the ones with the most stylists or the most generous terms. They are the ones whose POS was built to treat a hold and a consignment order as first-class states of inventory.
Key takeaway: A generic POS has no concept of a piece that is spoken for but not sold. Everything that tracks who has what, until when, and on what terms ends up in a spreadsheet or an associate's memory, and that is what fails when the program grows past one boutique. Stock commitment, expiry, deposits and card on file, one-step conversion and reporting have to be native to the POS, not bolted on.
Try Before You Buy Has Moved to the Mainstream
The clients who ask for approvals are the clients a house can least afford to inconvenience. Altagamma's Consumer and Retail Insight 2024 puts very important clients at around 30% of luxury brands' revenue on average. Those are the clients who are sent a selection before a collection lands, who shop with a stylist rather than a search bar, and who will not tolerate a hold that was quietly sold to someone else.
Reserve orders serve a broader group. A shopper who finds a piece online and wants to see it in person before committing will reserve it if the boutique lets them, and will travel to collect it if they trust that it will be waiting. That visit is the moment an associate can do what no website can: style the piece, add the accessory, open the client book.
Who is offering it
What was once the preserve of a flagship's private salon is now being offered by:
- Luxury houses sending curated selections to top clients on approval, with a stylist following up
- Jewelers and watch retailers holding a serialized piece for a named client, with a deposit, until they can visit
- Premium apparel and footwear retailers letting shoppers reserve a size online and collect it in store
- Multi-brand boutiques where a client's monthly edit leaves the shop on approval as a matter of routine
The commercial logic is sound. A hold brings a high-intent shopper into the store. An approval lets a client pay only for what they keep, which is a better experience than buying everything and returning most of it. Both deepen the relationship with the clients who matter most.
The operational logic is harder. And that is where most programs stall.
The Fundamental Mismatch: Why Generic POS Breaks Under Holds and Approvals
Every generic retail POS is built on a single assumption about inventory: a unit is available until it is sold. Available-to-sell is on-hand minus sold. Revenue is recognized when the card is tapped. Stock leaves the building only when it has been paid for.
Holds and approvals break every one of those assumptions.
A reserve order takes a unit out of available-to-sell without a sale. It has an owner, the client it is held for, and an expiry, after which it must return to sellable stock without anyone remembering to do it. A consignment order sends stock out of the building with no revenue recognized, against a deposit or a card kept on file. When the client decides, part of that order becomes a sale and part of it becomes stock again, and the two halves have to be reconciled on the same record.
A generic POS has no state for "spoken for but not sold." There is no field for who a unit is held for, no clock on the hold, no link between the pieces that left the boutique and the client who has them, and no way to turn two of four into a receipt while the other two go back on the floor. Everything that connects a piece to a client and a date has to be kept outside the system, typically in a spreadsheet or a stylist's notebook that becomes the most critical and most fragile record in the store.
Where the cracks appear
In one boutique with two stylists, manual oversight is manageable. Across a fleet, with reservations arriving from the website and approvals leaving from several stores, it fractures. The failure points are predictable:
| Workflow | What should happen | What generic POS forces |
|---|---|---|
| Reservation placed online | Unit comes off available-to-sell instantly, in every channel | Website and store both still show it as sellable |
| Associate notification | Push notification to pull the piece from the floor | An email someone has to see, or a phone call |
| Hold expiry | Uncollected piece returns to stock on its own | Someone checks a list, or the piece sits in the back for weeks |
| Selection sent on approval | Order records every piece, the client, the deposit or card on file | Manual stock adjustment and a note in a spreadsheet |
| Client keeps two of four | Two convert to a sale, two return to stock, one step | A fresh sale keyed by hand, two manual adjustments back in |
| Multi-store visibility | Any associate sees what is held and what is out, fleet-wide | A call to the other store |
| Director's view | Approval-to-purchase conversion and follow-up times on a report | Nobody knows |
The cost compounds quietly. Every piece held outside the system is a piece the website can oversell. Every approval tracked in a spreadsheet is stock that no count will reconcile until it comes back, if it comes back. Every manual adjustment is a shrink line waiting to happen. None of that is visible on a sales report, which is why it is tolerated for so long.
The workaround trap
Most retailers who start offering holds and approvals on a generic POS end up with a layer of workarounds: a reservations spreadsheet, a "pending" ticket type that nobody closes, stock adjustments out and back in, a stylist who carries the whole picture in her head. Each workaround adds administrative time, introduces a new error source, and consumes the hours that should go to the client.
The inflection point is not sales volume. It is the number of pieces that are spoken for at any one moment. The programs that scale smoothly are the ones that put the system in place before the growth arrived. The ones built on workarounds discover their limits at the worst possible time: a collection launch, a second country, or the day the stylist who knew where everything was hands in her notice.
What Scalable Holds and Approvals Infrastructure Actually Requires
Purpose-built hold and consignment logic is not a feature add-on. It is an architectural requirement, because it changes what "available" means for every channel at once. The distinction matters: a POS that can print a "reserved" ticket handles the simple case and collapses the moment the website, a second store and a deposit are involved.
Scalable infrastructure treats five capabilities as native.
1. Stock committed at the moment of the hold
Whether a client reserves online, by phone or through an associate, the unit must leave available-to-sell the instant the hold is placed, and every channel must read the same figure. This is not a flag on a ticket; it is the stock record itself changing state, so the website cannot sell what the boutique is holding.
2. Notification and expiry without human memory
The associate on the floor should get a push notification to pull the piece, not an email to a shared inbox. And the hold needs a clock. Whether the window is 24 hours, 48 hours or three trading hours, an uncollected piece must return to sellable stock on its own. A hold that depends on someone checking a list is a hold that will eventually be forgotten in the stockroom.
3. Consignment orders with a deposit or a card on file
A selection leaving the boutique on approval needs an order that records every piece, the client it went to, and the security behind it: an optional deposit, or a card token kept on file. Stock that has left the building without a sale is not an adjustment. It is an order in a specific state, and the system should know it as one.
4. One step from approval to receipt, and back to stock
When the client decides, the kept pieces convert to a completed sale on the POS and the rest return to the store's sellable stock, in the same action, with no manual adjustment on either side. This is where shrink is prevented or created. If "two kept, two returned" takes a new sale plus two adjustments, the count will drift and nobody will know which store it drifted in.
5. Reporting the retail director can act on
Approval-to-purchase conversion, how quickly associates follow up, and the client's purchase history should sit on the same record, so the director can see which stylists convert, which clients keep what they are sent, and where the program is paying for itself. A program nobody can measure is a program that gets cut at the first budget review.
The test for any POS: Can an associate in any store tell you, in under 30 seconds, which pieces are held for whom and until when, which pieces are out on approval with which client, and what is actually sellable right now? If the answer involves a spreadsheet or a call to another store, the infrastructure is not ready for holds and approvals at scale.
The Strategic Case for Getting This Right Now
The window for differentiation is open. Most luxury clients can reserve a table or a fitting online. Fewer can reserve the piece itself and know it will be there. Fewer still can have a selection sent home and pay only for what they keep, without a stylist improvising the paperwork. The houses that offer this cleanly are building the kind of loyalty that survives a price increase.
The cost of getting it wrong is relational as much as operational. A client whose hold was sold from under them, or whose deposit took three weeks to unwind, does not typically give a second chance. They take their business, and their top-client spend, to the house that got it right.
The broader integration question
For a house with more than one store, holds and approvals do not exist in isolation. A reservation placed on the website has to commit stock in the store the client will visit. A piece sent on approval from Milan has to show as out with the client when a Paris associate looks it up. A return has to reverse the right order, not just the store's revenue. This is why the capability belongs in the conversation about order management and inventory, not just the register.
It is also how Teamwork Commerce approaches it. Reserve orders come off available-to-sell the moment they are placed, the associate is notified to pull the piece, and a hold that is not collected within the window you set releases itself. Consignment orders carry an optional deposit or a card on file; kept pieces convert to a sale on the POS and the rest return to stock without a manual adjustment. The detail is on the luxury, jewelry and watches page.
What to evaluate before scaling
Before extending holds and approvals beyond one boutique, pressure-test the current POS against three questions:
- Does a hold change the stock figure every channel reads, or does it only print a ticket?
- Can a selection leave the store as an order, with a deposit or card on file, and come back as one receipt and a stock return in a single step?
- Does the picture hold across stores and channels, so any associate can see what is held, what is out, and what is sellable, without a phone call?
If any answer is "no" or "not exactly," the infrastructure is not ready for what the best clients now expect. The retailers building holds and approvals on a POS that treats them as native are not just fixing an operational problem. They are turning the most expensive moment in luxury retail, the one where a client hesitates, into the moment the relationship deepens.