ShipStation Now Applies Packing Recommendations Automatically

Short answer: Since October 2026, Packing Recommendations is an automation rule action called Recommend Package. Add the rule and ShipStation computes a packing plan from your product dimensions and configured packages and applies it at import, multi-package configurations included, without anyone opening the order. Rate Shopper then prices the package it produced. It needs dimensions on every product and a maximum weight on every package, it solves for the fewest packages that hold the order rather than the cheapest, and because it is a rule, where you put it in the list decides whether it wins or gets overwritten.

This is a real improvement, and most operations with a varied catalogue should turn it on. The rest of this page is what it does, what it needs, and the five things worth checking afterwards.

ShipStation has not documented the change yet, so everything below is from accounts I work in rather than from the help centre. The Packing Recommendations article still describes the older per-order flow and remains correct on the prerequisites.

What changed

Three things, and the third is the one that matters most.

It is a rule action. Recommend Package appears alongside Set Package Dimensions and Set Rate Shopper in the actions list. That means it runs on the orders your rule’s criteria match, at import, like every other rule.

It applies the plan. Previously you opened an order, asked for a recommendation, and applied it. Now the plan lands on the order, and where the order needs more than one box, the multi-package configuration is built.

The rate prices the result. An automation rule assigns the Rate Shopper configuration at import; the pricing itself runs whenever the rate is requested and again immediately before the label is bought, against whatever package is on the order at that point. So the box that Recommend Package chose is the box that gets priced.

That closes the gap this site has described for a year: that packing and rating were one decision being taken as two, because the packing half happened after the rating half and only when somebody asked for it. That is no longer the case.

Where it sits, and why that decides whether it works

Automation rules run last, after Auto-Routing, Auto-Split, SmartFill, Preset Groups, Product Defaults and Service Mapping. Within the rules list, ShipStation processes top to bottom and later rules override earlier ones.

Recommend Package is subject to both. So:

Put it above your Rate Shopper rule. ShipStation already documents this dependency for the dimensions rule — a rule the Rate Shopper rule depends on has to run before it. The same reasoning applies here, because Rate Shopper prices the package it is handed.

Anything that sets a package below it overrides it. An item-count rule, a Set Carrier/Service/Package action, a Preset Group value reasserted by a rule — if it sits lower in the list, it wins.

And anything above it gets overwritten. This is the one to check first, below.

Five things to check in your own account

1. Your cold-chain or special-handling rules

If you have a rule that applies an insulated box, a foam-lined carton, or any packaging chosen for the product rather than for fit, and Recommend Package sits below it in the list, your rule is being replaced on every matching order. Nothing warns you. The order still ships; it ships in the box the packing plan chose.

Move the special-handling rule below Recommend Package, or scope Recommend Package’s criteria to exclude those orders.

2. Product dimensions

Nothing has changed here, except that it matters more. Packing Recommendations will not run without per-SKU dimensions. A catalogue without them does not get a degraded plan; it gets nothing. Length, width and height as packed, not as marketed, on the SKUs that carry your volume.

3. Your package list

Every box you actually stock, with real internal dimensions and a maximum weight. Internal, not external — the difference is the wall thickness, and it is the difference between a box that closes and one that does not. The plan can only pack into boxes it has been told about. Interior is what the plan packs into, not what the carrier bills: once the plan has run, bulk-update those orders to the box’s true exterior dimensions before buying labels. ShipStation does not swap one set for the other, so that step is manual.

4. Orders where the service is already pinned

USPS and FedEx One Rate do not support multi-package labels. If Rate Shopper chooses the service, this does not come up: it will not select a service that cannot carry a multi-package label. But where the service is set before that — by Service Mapping from the customer’s checkout choice, by a static rule, or by hand at the bench — an order can reach the label with a multi-package configuration its service does not support, and label creation fails.

Worth checking against how your services actually get set rather than assuming.

5. What your packers would have done

For the first week, compare a sample of plans against what your bench would have chosen. Mostly they will agree. Where they don’t, the reason is usually in the next section, and it is worth knowing which of your SKUs produce the disagreement.

What a packing plan reads, and what it can’t

A packing plan works from what the product and package records hold: dimensions, weights, a maximum weight. That is enough for boxed goods of a predictable shape. Three common cases sit outside it, and they sit outside it because the records have nowhere to put the fact, not because the calculation is wrong.

Contents that give. A bag of coffee beans, a soft-sided pouch, a garment. A packing model treats each as a rigid block of its stated dimensions. Your packers know three bags fit a carton that the numbers say holds two.

Contents with usable interior space. A bin has a volume inside it. A bottle leaves room around its neck. A model reading only outside dimensions counts both as solid.

Packaging whose interior changes. A ShipStation package record holds one set of dimensions, and there is no way to say that a carton is used with an insulated liner in July and without one in January. One record, two usable interiors, and whatever reads it is working from the summer answer in winter or the winter answer in summer.

The same applies to weight. A package record does carry a tare weight, but only one, and a carton with an insulated liner weighs more than the same carton without it. Whichever figure is on the record is right for half the year. Where the parcel is weighed after it is packed, the scale settles it; where the weight is computed from the order instead, a lined carton can be declared at its unlined weight.

None of this is a defect. A horizontal platform serving a very large and varied merchant base cannot reasonably model which of your products compress, which of your containers have interiors, or which of your cartons take a liner in summer. Those are facts about one operation.

What this does not change

It solves for fit. It returns the fewest packages that hold the order. That is the right answer most of the time and it is not the same question as which packaging costs least to send.

Rate Shopper compares services, not packages. It prices the box it is given, across the services you made eligible. The box is the one input it never varies.

So the rate follows the box: change the box and it re-prices against it. What does not happen is the reverse. The box is never varied as a candidate, which is where carrier flat-rate programs sit, and where a cubic tier boundary your cartons sit close to sits.

If you are setting this up

Dimensions on your top SKUs by volume first. Packages configured with real internal dimensions and a maximum weight, and a bulk update to exterior dimensions before labels are bought. Then add Recommend Package, above your Rate Shopper rule and above nothing that needs to survive it. Then watch a week of orders against what your bench would have done.

For most operations that is the whole job, and it is a better answer than ShipStation had a month ago.