Supplier Sourcing, Stage by Stage

Supplier sourcing means finding, qualifying, and contracting the companies that make or supply your product. For a small brand that's four stages: define the spec, build your candidate list from multiple channels, qualify with samples and written terms, then contract and set up your reorder rhythm. Most founders jump straight to stage two.

That is how you end up qualifying factories against a product you never defined. Skip the spec and you end up sourcing against a half-formed idea. Skip multiple channels and you're seeing a slice of one marketplace's algorithm, not the actual market. Rush qualification and you land a supplier you can't replace, on terms you never negotiated, for a product you never fully defined. This guide walks each stage in order and shows what it costs in time.

Stage one: write the spec before you look for anyone

You cannot qualify suppliers against a product that doesn't exist on paper. A spec doesn't need to be an engineering document. It needs to be specific enough that two different factories quoting it would produce something comparable.

Minimum: physical format and size, materials or ingredients, packaging format, target landed cost per unit, quantity for the first run, date you need stock in hand. Six lines. That's the whole spec.

The landed cost line is what founders skip and then regret. Landed cost is the unit price plus freight, duty, insurance, and inspection. On typical imported goods that's 30 to 50 percent more than the quoted price. On bulky or low-value items it can approach double. If you quote suppliers on ex-works price alone, you're hiding the number that actually controls your margin.

Stage two: build the candidate list from at least two channels

One marketplace is not a list. It's a slice of one ranking algorithm. Use two channels minimum. Three is better.

Sourcing marketplaces get you breadth fast. Alibaba dominates for overseas manufacturing. Faire covers finished wholesale goods. For print on demand there's Printful, Printify, and Gelato, none of which carry minimums, which makes them the cheapest way to test a design before committing to a run.

Domestic directories return fewer listings, but a far higher share of them are real factories. ThomasNet for North American industrial suppliers. Maker's Row for apparel and leather goods. Fewer results. Much higher average quality.

Trade shows give you qualification for free, because exhibiting filters out the least serious players. One day walking a show usually beats a week of catalogue browsing.

Reverse sourcing is the move nobody uses. Find a product you admire, read the packaging for a manufacturer mark or fill location, approach that facility. You start with proof they already make the category.

Aim for eight to twelve names. Fewer than eight and you won't get enough replies to compare. More than twelve and qualification becomes a bottleneck.

Stage three: qualify with samples and written terms

Qualification is where sourcing is won or lost. Budget four to six weeks and don't rush it.

Order samples from three suppliers, not one. One sample shows you what a factory can do on its best day. Three show you what normal looks like in your category and price band. Pay for the samples. A supplier offering them free at any volume has priced that cost back into your unit rate.

Get pricing at three quantities, not one. The shape of the price curve tells you where the factory wants to operate. A supplier whose price barely moves between 500 and 5,000 units is probably a middleman, not the producer.

Ask what happens when a run comes back defective. Who pays for the batch? Who pays return freight? What defect rate counts as failure? Get the threshold as a number, in writing. A supplier who won't answer in writing has answered.

Verify certifications against the issuing body's own register, not the PDF they sent you. For anything ingestible or applied to skin this is standard business, not an accusation.

Our full walkthrough for vetting a factory, including the exact outreach email that gets replies, is in the guide to finding a private label manufacturer.

Stage four: contract, then set the reorder rhythm

The contract stage is quick if stage three was done right. Confirm unit price and price breaks, MOQ, lead time for first and repeat runs, payment terms, the defect policy, and who owns tooling if you paid for any.

Payment terms for a new relationship are usually 30 percent deposit and 70 percent on shipping documents. Paying 100 percent up front to a supplier you've never worked with is the single most common way small brands lose money in sourcing.

Then set the reorder rhythm immediately, before you need it. Your reorder trigger is the stock you will sell during the lead time, so work it out in units rather than in vague weeks of cover.

Say you sell 50 units a week and a repeat run takes 8 weeks door to door. Your trigger is 50 times 8, so 400 units. If your first delivery was 500 units, you reorder once 100 have sold, roughly two weeks in.

Reordering at 100 units remaining instead would leave you out of stock for six weeks. Most first stockouts are arithmetic failures rather than demand surprises, and that multiplication is the whole of the arithmetic.

Overseas or domestic: the trade you are actually making

The overseas versus domestic question gets argued on unit price, which is the least interesting variable. The real trade is between unit cost and cycle time. Cycle time is what constrains a small brand.

An overseas supplier at 40 percent lower unit cost with a twelve week lead time is worse than a domestic supplier at full price with a three week lead time, if you're still learning what sells. Short cycles let you reorder the winners and abandon the losers. Long cycles force you to bet the whole season up front.

The rule that holds: source domestically while you're discovering demand. Move overseas once your reorder pattern is predictable and volume makes the freight math work. Plenty of brands never make the second move and do fine.

If you're selling before you hold stock at all, dropshipping removes the sourcing timeline entirely at the cost of margin and brand control. That's a reasonable trade during validation.

Single sourcing, and when it stops being fine

Almost every small brand single sources early on, and that's reasonable. One supplier means better volume pricing, a simpler relationship, and consistent output.

It stops being reasonable the moment the supplier becomes hard to replace. A factory holding your tooling, your formula, and your only qualified production line has pricing power over you. And it will use it.

Cheap insurance: qualify a second supplier to sample stage even if you never order from them. That costs a few hundred dollars and a few weeks now, against a multi-month scramble during a stockout later.

What sourcing costs in time, honestly

From written spec to delivered first run, plan on three to six months for an overseas supplier. A domestic route at two to five week lead times finishes in well under three months. About one week of that is list building. Four to six weeks is qualification. The rest is production and freight.

Founders underestimate this by half. Then they compress the qualification stage to catch up, which is the stage that most needs the time. If your launch date is fixed, move the date or narrow the product. Don't shorten qualification.

Our teardown of what it costs to start a clothing brand walks the same timeline with category-specific numbers.

Once stock lands, the operational load shifts to orders, refunds, and returns. Our guide to the ecommerce back office covers that handoff end to end.

Where an AI co-founder fits

Sourcing is one of the functions SoGood's Operations department runs directly on the Expert tier. That covers supplier shortlisting, outreach correspondence, and fulfillment coordination once a supplier is chosen. It keeps running after launch rather than stopping at the store build.

SoGood is priced in tiers: Basic is free, Pro is $29 a month, and Expert is $99 a month, and you can add credit packs on any plan. Sourcing sits on the Expert tier, so compare it against that figure.

The limits are real. It doesn't negotiate or sign contracts. It doesn't handle customs, duty, or compliance paperwork. It can't tell you whether a sample feels right. What it does: compress the search and the correspondence, which is most of the calendar time. The judgment calls stay yours.

For how sourcing connects to the rest of the operation, see our piece on whether AI can source suppliers and handle fulfillment.

What to do this week

Write the six-line spec today, including a landed cost target rather than a unit price target. One hour, and every quote that comes back is finally comparable.

Then build the candidate list from two channels and send outreach to all of them at once, not sequentially. The four to six weeks in the table assumes you contacted everyone at once. Run sequentially and that stage stretches to months. Run outreach sequentially instead and that one stage stretches into months.