Shopify Flow can now buy shipping labels by itself. Sort your orders before you switch it on

What Flow can do now

On October 15, 2026, Shopify added an action to Flow, its built-in automation tool, that buys shipping labels on its own. You set conditions, such as an order tag, specific products, single-item orders, or any combination of order details. You pick the carrier, service, package, weight and ship date. When an order matches, Flow buys the label and it's ready to print. Orders that don't match are left alone for you to handle as usual. Shopify provides a template to start from.

For a store shipping dozens of near-identical orders a day, this removes a lot of clicking. It also moves a decision, which orders get a label and with what settings, from a person looking at each order to a rule that runs without anyone looking.

That's fine for most orders. It isn't fine for all of them. A label bought for an order you then cancel, change or hold for review is money spent and a task to undo. Automation is easy to switch on. The work is deciding which orders it should never touch.

So before you build the workflow, sort your orders into three piles.

Pile one: automate

These are the orders where the right label is obvious and the same every time.

Single-item orders of a known product. One candle, one book, one T-shirt. You know the package and weight, and you know the service you use for it.

Standard domestic addresses. Orders shipping within your country, to an address the customer entered at checkout and nobody has touched since.

Paid, unremarkable orders. Fully paid, not flagged by your fraud checks, no notes from the customer asking for anything unusual.

An illustrative first rule might be: single-item orders of your three best-selling products, shipping domestically, with no customer note and no hold, get your standard service in the usual box. That's narrow on purpose. You can widen it once it's run smoothly for a couple of weeks.

Even in this pile, keep the ship date sensible. If orders come in overnight and you ship the next afternoon, set the date to match how your team actually works. A label dated for a day you aren't shipping causes confusion at the counter and with tracking emails to customers.

Pile two: look first

These orders can often go through automatically, but a person should glance at them first, at least until you trust the rule.

Multi-item orders. Weight and package size change with the combination. If you've set up Flow with one package and one weight, a three-item order will get the wrong label.

Higher-value orders. You may want a different service, insurance or a signature, and a person is better placed to decide.

Orders with a customer note. Notes are where people ask for gift wrap, a delay or a different address. Flow can check whether a note exists, but it won't understand what it says.

Repeat customers with special arrangements. Wholesale buyers or regulars who always get a certain service.

The aim of this pile is to shrink over time. Each week, look at which look-first orders turned out to need nothing changed. If a type of order always goes through untouched, it probably belongs in pile one, with its own rule and its own package settings.

One practical approach is to tag these orders automatically, for example with a "check label" tag, so they land in a short list for a person to review each morning.

Pile three: never automatic

Some orders should always get a human decision, however good the rule looks.

Orders on hold for fraud review. If your fraud checks or your team have put an order on hold, a label shouldn't exist until someone clears it. Make sure your Flow conditions exclude these explicitly rather than relying on timing.

Orders with edited addresses. If a customer emailed to change the address, or your team edited it, check it before a label is bought. A wrong address on a printed label is the most common way automatic shipping goes wrong.

Restricted items. Perfume, batteries, aerosols and other items carriers restrict need the right service and sometimes special labeling. Our note on why perfume can't fly explains the trap. Keep these manual.

Pre-orders and backorders. A label bought for something you can't ship yet sits there waiting, and the ship date on it will be wrong.

Local pickup and local delivery. These shouldn't get a carrier label at all, but a loose rule can catch them.

International orders, at least at first. Customs details and service choices vary too much for a first rule.

Before you switch it on

Check three things in your setup.

Your product weights. Flow will use what you set, but if your product records have wrong or missing weights, it'll make the same mistake on every order. Fix them first.

Your exclusions. Write the pile-three conditions into the workflow as explicit exclusions, so the rule says what not to touch as well as what to touch. Test it against a handful of recent orders, including ones from each pile, and confirm only the right ones would match.

What happens when something goes wrong. Decide who checks the printed labels each morning and what they do with a label that shouldn't exist. Find out from your carrier and from Shopify how voiding or refunding a label works for your account before you need it, not on the day.

Start the rule on one shipping day, watch it closely, then let it run for a week before widening it. Busy weeks like Black Friday are a bad time to switch on something new, so if you're going to do this before the holidays, do it soon.

Put this into practice

This week: list your ten most common order types and put each in one of the three piles. Next: fix product weights for everything in pile one and build a narrow rule with explicit exclusions. Then: test it against recent orders, run it for a day with someone watching, and widen it slowly.

If you want shipping, inventory and order handling automated across Shopify, your warehouse and the other tools you use, with the exceptions handled properly, that's part of our business automation work. Talk to us.

Related work: Paddle8.