A rare-event process is the one that needs an SOP most. Six phases, five owners, one graduation gate.
A DTC brand onboards a new supplier maybe twice a year. A new co-packer after a quality scare, a second packaging vendor before Q4, a cheaper manufacturer once volume finally justifies the switch. Rare enough that nobody owns the process, consequential enough that improvising it costs real money.
Improvised onboarding looks like this: the founder negotiates terms over email, the ops lead requests a sample, nobody asks for the insurance certificate, and the lead time that lands in the reorder sheet is whatever number the sales rep quoted. Six months later the brand is reordering off fiction and accounts payable is paying invoices against terms nobody wrote down.
This post is the onboarding SOP: six phases, five owners, and the 90-day review that tells you whether the supplier you onboarded is the supplier you thought you were getting.
Frequency is the problem. Your team processes refunds daily and ships orders hourly, so those processes get documented, corrected, and re-corrected until they work. Supplier onboarding happens a handful of times a year, months apart, usually handled by whoever has capacity that week. There is no repetition to sand the process smooth.
So each onboarding starts from scratch. The person running it reconstructs the last one from an email thread, misses the two steps they never knew existed, and invents the rest. The knowledge from every previous onboarding, the question that exposed the bad co-packer, the clause that saved the Q4 order, lives in one person's head or nowhere.
Rare-event processes need documentation more than daily ones. Not because the steps are complicated, but because there is no muscle memory to fall back on.
The failures are specific, and they repeat across brands:
Six phases, in order. Each phase has a named owner and an exit condition, and a supplier does not advance until the exit condition is met. The relationship is not live until the test PO passes.
Three quotes minimum, then vet the suppliers behind the best two. What you are actually checking: can they handle your Q4 volume, not just your January volume. Will they exist next year (ask for references from two brands roughly your size, and call them). Do their certifications match your category's requirements. And how they behave now is the ceiling: a supplier who is slow and vague during the sales process will be slower and vaguer once they have your deposit.
Exit condition: a completed vetting sheet and two reference calls logged.
Request a production sample, made with the real materials on the real equipment, not a hand-finished prototype. Review it against a written spec: materials, dimensions, weight, color, packaging, labeling. When it passes, both parties sign off in writing, you photograph it from every angle, and each side keeps one physical unit.
The golden sample is the contract that settles every future quality dispute. When a production run arrives wrong, the conversation is 'this does not match the golden sample,' with photos attached, instead of an argument about expectations.
The inspection procedure the golden sample anchors: what to check on every inbound run and who signs off.
Exit condition: a signed golden sample, photographed, one unit with you and one at the supplier.
Negotiate everything into one written terms sheet: unit pricing with quantity breaks, minimum order quantity, payment terms, and who pays for which leg of shipping. On payment terms, 30 percent deposit with the balance before shipment is standard for overseas manufacturing; net 30 is achievable domestically once trust exists. On shipping, the Incoterm decides whether the quoted price includes freight, insurance, and customs, and a cheap EXW quote regularly loses to a pricier DDP quote once you pay for everything in the middle.
Then build the real lead time. Take the quoted production time and add sampling rounds, pre-shipment inspection, freight for your actual shipping mode, customs clearance, and receiving time at your 3PL. That total, not the quote, goes into your inventory tool, because it is the number your reorder points depend on.
Why reorder points built on the supplier's quoted lead time run you out of stock, and the weekly routine that keeps cover honest.
Exit condition: a signed terms sheet and a documented total lead time.
Collect the paperwork before the first PO, because you have no leverage after it. What to collect depends on what you sell:
| Category | Collect before the first PO |
|---|---|
| Every category | Certificate of liability insurance naming your company, W-9 (domestic) or W-8BEN-E (overseas), signed terms sheet |
| Supplements, food & beverage | FDA facility registration, GMP certificate, certificate of analysis per lot |
| Cosmetics & personal care | MoCRA facility registration, ingredient documentation, stability testing |
| Children's products | Children's Product Certificate backed by third-party CPSIA lab testing |
| Electronics | FCC certification, UL or equivalent safety listing, battery documentation for freight |
File everything where the next person can find it, and calendar the expiry dates. An expired insurance certificate is the same as no certificate.
Exit condition: the category checklist complete, with expiry dates calendared.
This is the phase improvised onboarding skips, and it is where the cost compounds for years. Everything negotiated in phases 2 through 4 has to land in the systems that run the business, or it might as well not exist:
Where the payment terms you negotiated get enforced, or silently ignored: the AP three-way match inside the monthly close.
Exit condition: a second person confirms each system shows the negotiated values. Not the person who entered them.
Place a small real order and run it through the entire path: PO issued, production, third-party pre-shipment inspection if overseas (an AQL inspection costs around $300 and is the cheapest insurance in this SOP), freight, customs, 3PL receiving, stock going sellable. Grade every hop. Did production match the golden sample? Did the paperwork clear customs without a hold? Did the 3PL receive without exceptions? Did the invoice match the PO?
Do not cut the incumbent until the new supplier has passed the test PO and one full production cycle. Overlapping costs you one cycle of double stock on one SKU. A failed first run with no fallback costs you a stockout on every SKU that supplier makes.
Exit condition: test PO received, graded, and reviewed. Only then does the supplier get real volume.
Onboarding fails at handoffs, so name them. Five roles; on a small team one person holds several, but each is a name, not a department:
Ninety days after the first real PO, score the supplier on four measures: on-time delivery against confirmed dates, defect rate from inspections and customer returns, invoice accuracy against POs, and responsiveness when something went wrong. Score each 0 to 2 and write down the total.
The score has two jobs. It decides whether this supplier earns more volume, keeps current volume, or triggers a backup search. And it becomes the baseline the next quarterly review compares against, which is how a slow slide in performance gets caught in one quarter instead of discovered during a Q4 crisis.
The same score-then-review structure applied to your fulfillment partner, with the SLA numbers worth writing into the agreement.
Everything you captured during onboarding starts aging the day the test PO clears. Your contact leaves the supplier and the escalation path goes dead. The lead time stretches from 45 days to 65 across two busy quarters. A price break threshold moves. Payment terms change in a renegotiation nobody mentioned to finance.
This is SOP drift applied to supplier management. The record looks complete precisely because someone did the work once, and complete-looking data is the kind nobody re-checks. Two habits contain it: when any PO arrives late, the ops owner updates the lead time in the supplier record and the reorder points that depend on it, that day. And the quarterly scorecard review re-verifies the record's load-bearing fields: contacts, terms, lead time.
The reason to make this an SOP rather than a one-time checklist is the gap between runs. A process that fires a few times a year, with months of silence in between, is where steps get skipped. In ReccordSOP, the onboarding SOP gets dispatched as a checklist run for each new supplier: every phase is checked off by its owner, the completed run is a permanent record of exactly what was verified for that specific supplier, and the screen-heavy phases, the systems setup and the PO raising, are recorded once and regenerated as screenshot steps when your tools change.
Onboard your next supplier against a written SOP instead of an email thread. Generate the systems-setup steps free at reccordsop.com.
Four to six weeks domestically and eight to twelve overseas, measured from first contact to a passed test PO. Sampling rounds move the timeline most: each golden sample revision adds two to three weeks, which is why a written spec before sampling pays for itself.
A production-grade sample, made with the real materials on the real equipment, that both you and the supplier sign off on and physically keep. It becomes the standard every future production run is judged against, which turns quality disputes into comparisons instead of negotiations.
At minimum: a certificate of liability insurance naming your company, tax paperwork (W-9 domestically, W-8BEN-E overseas), and a signed terms sheet. Then category-specific compliance: FDA registration and per-lot certificates of analysis for ingestibles, MoCRA documentation for cosmetics, a CPSIA-backed Children's Product Certificate for kids' products, FCC and safety listings for electronics.
Always, and make it a real order that travels the full path from PO to sellable stock at your 3PL, not a sample shipment to the office. The test PO is the only step that checks the supplier's paperwork, packaging, and logistics rather than just the product, and those are what fail at customs and receiving.
Run a 90-day scorecard on four measures: on-time delivery, defect rate, invoice accuracy, and responsiveness, each scored 0 to 2. Repeat it quarterly against the baseline. Volume decisions then follow the numbers instead of the relationship's founding optimism.
I built ReccordSOP after watching too many DTC ops teams lose months to undocumented workflows. These SOPs are battle-tested with Shopify operators running $1M to $50M brands.
Last reviewed August 4, 2026
Most onboarding guides are written by 3PLs for their own benefit. This is the merchant's playbook, and the scorecard that holds any partner accountable.
Stop reordering by vibes. A weekly, owner-assigned process that keeps your bestsellers in stock without freezing cash in dead inventory.
Shopify made B2B a toggle. The operations are what break. Here are the SOPs to write before your first wholesale order ships.
Most SOPs are wrong within 90 days of publishing. Here's how to detect it before it costs you a customer.
We use essential cookies for sign-in and a small amount of analytics to improve the product. Privacy policy.