Yes — you can now put a subscription product on a Shopify draft order, email the invoice, and have a real subscription contract created when the customer pays. Shopify switched this on at the end of October 2025. Before that, a shopify draft order subscription was a contradiction: drop a product with a selling plan into a draft order and it came through as a plain one-time purchase, no contract, no recurring billing, nothing. Sales reps worked around it with prefilled cart permalinks, manual contract creation inside Recharge, or a developer-built checkout that nobody wanted to maintain.
That workaround tax is gone for most stores. Which matters more than it sounds, because the orders that get placed over the phone — office coffee, clinic backbar, kennel and groomer pet food, wholesale replenishment — are exactly the orders that should have been recurring all along.
What actually happens when the customer pays the invoice
The flow is short. In the admin you create a draft order, add a product that has a selling plan attached, pick the plan (every 30 days, every 2 weeks, whatever the plan group offers), set shipping and the customer, then send the invoice. The customer gets an email with a checkout link. They enter a card. On successful payment, Shopify creates the first order and a subscription contract, with the payment method vaulted for future cycles.
The billing anchor is the date they paid, not the date you built the draft. So if your rep writes the order on a Tuesday and the buyer sits on the invoice until the following Monday, every subsequent charge lands on a Monday. For monthly plans nobody cares. For a weekly coffee delivery routed around a Friday office close, somebody will care, and you'll be editing the contract afterwards.
The part that breaks: "mark as paid" does not create a subscription
Recurring billing needs a stored payment method. There is no way around that. If you collect a check, take a wire, or hit Mark as paid because the buyer gave you a PO number, Shopify has nothing to charge next month. You get an order. You do not get a working contract.
This is the single most common mistake we see when a team first turns this on, and it's the awkward one for B2B, because net-30 buyers are the people most likely to want replenishment. Two honest options:
- Get a card on file anyway and treat it as the instrument of record, with terms handled separately in your accounting. Plenty of mid-size accounts will do this for a $400 monthly order. Very few will do it for a $40,000 one.
- Keep large net-terms accounts out of subscriptions entirely and run them as scheduled draft orders your rep reissues, which is manual but at least it's honest about what it is.
Shopify's B2B payment terms and its subscription billing engine are separate systems. Don't design a program that assumes they'll meet in the middle.
B2B subscription setup on Shopify, without the custom checkout
On Plus, B2B company profiles, locations and catalogs were already the right place to hold negotiated pricing. What changed is that the sales-assisted path now reaches the subscription engine, so a rep can build a recurring order against a company location instead of handing the account to a developer.
One trap worth naming: the discount you type into a draft order applies to that draft order. The recurring price is governed by the selling plan. If you've promised a 22% standing discount to a distributor, build that as a selling plan in a plan group tied to their catalog — don't type it into the line item and assume it carries. Test it by creating a plan on a $1 test product, paying the invoice with a real card, then using Shopify's billing attempt tooling or simply waiting out a one-day interval to watch what the second charge actually is. We get this wrong on the first pass more often than we'd like, usually because the plan group was published to the wrong catalog.
Also: multi-location fulfilment. If the buyer wants the same plan shipped to four branch offices, that's four contracts, four shipping addresses, four sets of rates. Shopify won't split one contract across locations for you.
Corporate replenishment: the arithmetic that justifies the whole thing
Take a roaster selling to offices. An 80-person company buys 14 bags of whole bean at $21 a bag. Left to reorder manually, the office manager places maybe nine orders a year — they forget, they run out, someone buys from the grocery store down the street in the gap. That's 9 × 14 × $21 = $2,646 a year.
Put the same account on a 14-day plan and it's 26 deliveries: 26 × 14 × $21 = $7,644. Give them 10% for committing and it's $6,880. The account is worth 2.6× what it was, and the only thing that changed is who remembers to place the order.
Now the cost side. A rep spends about 20 minutes on the call and the draft order. At a $30/hour loaded rate that's $10 of labour to originate a $6,880 account. Compare that to what you're paying Meta for a cold prospecting conversion right now. The reason retention and subscription programs keep eating budget share in the US isn't that they're fashionable — it's that CAC on paid social has gone up enough to make a phone call look cheap.
Run your own version of that with your real reorder frequency before you build anything. If your product is genuinely replenished every four months, the multiple collapses and this is a smaller deal for you.
Phone orders for subscription products: what the rep says and records
Selling a recurring charge over the phone puts you in negative-option territory, and US state law is where the actual exposure sits. California's automatic renewal statute is the strictest and the most copied; several other states have their own versions. The practical requirements rhyme across all of them: disclose the amount, the interval and the fact that it renews, before the customer agrees; make cancellation genuinely easy; and keep evidence of consent.
So the rep's script needs four things said out loud, in this order: what ships, how often, how much per shipment, and how to cancel. Then the invoice email does the work of written confirmation, because the checkout page states the terms and the customer actively pays. That's a cleaner consent trail than a rep ticking a box in a CRM, which is a quiet argument for the invoice flow over keying a card in yourself.
Add a note field on the draft order where the rep logs the call date and who authorised. Boring. Saves you an afternoon when a chargeback lands.
Recharge and draft orders: do you still need the app?
Depends on what you're doing with it. Shopify's native subscription capability covers fixed intervals, prepaid and pay-as-you-go, discounts per plan, and customer-managed changes through the account area. If that's your whole program, the native route plus draft orders is now a complete answer and you can stop paying a percentage of subscription revenue to a third party.
Third-party apps still earn their keep on: build-a-box and swappable bundles, deep retention tooling (dunning sequences, cancellation flows with offers, win-backs), SMS-driven upcoming-order edits, and analytics your finance team actually asks for. If you run Recharge and want the draft-order path, check that their subscription products are exposed as selling plans Shopify recognises at the draft order stage, and test an end-to-end invoice before you train a sales team on it. App support for new Shopify surfaces arrives on the app's schedule, not Shopify's.
Migrating contracts between subscription systems is its own project — payment tokens, billing dates, remaining prepaid cycles. It is not a weekend. If you're weighing it, talk to someone who has done one; our app and integration team has unpicked enough of these to tell you quickly whether it's worth the disruption.
Sales tax on recurring shipments
Each billing cycle is a taxable event in its own right, calculated at the address at the time of shipment. Two things to watch. First, economic nexus: a subscription program that quietly adds twelve shipments a year per customer pushes you over state thresholds faster than one-time sales do, and you'll cross into a new registration without a single marketing campaign telling you about it. Second, rate drift — a customer who moves, or a jurisdiction that changes its rate mid-year, gets recalculated per cycle, so the revenue you forecast off cycle one won't match the remittance.
Shopify Tax handles the calculation for most catalogs. Avalara is the call when you have messy taxability (food and supplements are exempt in some states and not others, and that distinction is where coffee, snack and wellness subscriptions get bitten) or when you're filing in more than a handful of states.
Short answers to the things people ask next
Can I invoice a customer for a subscription in Shopify? Yes, that's the draft order invoice. Build the order, send it, they pay, the contract starts. What you cannot do is invoice on terms and have recurring billing work, for the reason above.
Does a subscription contract get created from a draft order automatically? On payment, yes. It appears under the customer record like any other contract and can be edited, paused or skipped from there. If no contract shows up after a paid invoice, the line item didn't carry a selling plan — the usual cause is that the product's plan group isn't published to the market or catalog the draft order was created in.
Can a rep take the card over the phone instead? Technically yes, through the draft order's payment collection. We'd rather you didn't. Reading card numbers down a line adds PCI scope you don't need, and the invoice link gives you the consent record for free.
Will this handle a corporate account that wants one invoice for four offices? No. Four contracts, four charges. Consolidated invoicing across subscription contracts is still a custom build.
Where this is worth your attention before BFCM
Gift subscriptions and concierge selling spike hard in November, and a rep who can close a recurring order on the call instead of mailing a cart link is the difference between a booked program and a maybe. If you're going to roll this out, do it in September, not the third week of November. Train two reps, not twelve. Ship 30 contracts, read the dunning behaviour on cycle two, then widen.
Pick one account type you already sell to by phone — offices, clinics, groomers, whoever reorders on a rhythm. Build one selling plan for them, run a single $1 test from draft order to second billing attempt, and watch exactly what the customer sees. If the second charge is the amount you expected, you have a program. If you want a second pair of eyes on the plan structure or the B2B catalog setup, our US team does a free store audit and this is a reasonable thing to point it at.