How to find a manufacturer for your product when you're already selling

Most "how to find a manufacturer" advice is written for people starting from zero. If you're reading this, you're probably not starting — you're re-sourcing a product that outgrew its supplier, splitting volume to reduce dependency, or adding a category and discovering your current factory doesn't make it. The question changes shape when you already sell: it's not "can anyone make this?" but "who can make this reliably, at your volume, for 40 more reorder cycles?" Reliability is the whole game, because your reorder math, your buying calendar, and your cash cycle are all downstream of one number: how long it actually takes this factory to deliver.

Start from your requirements, not their catalog

Before contacting anyone, write down what the new supplier has to hit. Not aspirations — constraints, pulled from your own history:

• Real lead time you need. If your current supplier's actual delivery time (not the quote — measure it from your PO history) is 45 days and your buying list is built around that, a new supplier quoting "35 days" needs to prove it, not say it.
• Volume profile. Your average reorder, your peak-season reorder, and how much those vary. A factory comfortable at 500 units and a factory comfortable at 5,000 units are different factories.
• Quality floor. The top two or three defects you currently see — stitch density on dog harnesses, plating wear on enamel pins, seal integrity on coffee bags. These become your acceptance criteria, and they should be in writing before the first quote.
• Compliance needs. Anything sold to kids, pets, skin, or food triggers testing requirements that vary by market — US entry and product-safety requirements are outlined on CBP's Basic Importing and Exporting page (verified October 2026). Confirm what applies to your category with a testing lab before you fall in love with a quote.

A supplier who can make your product is abundant. A supplier who can make your product at your volume, on schedule, with documented QC, is rare — and the search should filter for the second kind.

Where to actually look

Industry-specific directories and trade shows. For categories with real manufacturing specialization (pet gear, textiles, stationery), niche sourcing directories and regional trade shows surface factories that generic platforms bury (the Thomasnet directory, for example, indexes North American manufacturers by category — Thomasnet, accessed October 2026). You pay in travel time; you get suppliers whose entire business is your category.

Alibaba and similar B2B marketplaces. Fine for discovery, weak for verification — treat listings as leads, not facts. Filter for manufacturers (not trading companies) if vertical integration matters to you, and check how long the account has existed and whether the listed certifications are actually present as documents, not icons.

Your current supplier's network. Factories specialize and subcontract. Asking your existing supplier "who makes [adjacent product] well?" often produces the highest-quality lead available, because factories know each other's reputations and rarely recommend bad ones.

Domestic and nearshore options. For replenishment of fast movers, a factory with a 15-day domestic lead time at a higher unit cost can be worth more than a cheaper overseas one — the tradeoff shows up in your reorder math, not just the quote (more on that tradeoff in how much inventory to order).

Vetting for reliability: what actually predicts on-time delivery

This is where most re-sourcing searches fail, because the evidence people collect (photos, reviews, a nice sample) says nothing about schedule discipline. What actually predicts delivery:

1. PO history, not promises. Ask for two or three reference customers at roughly your volume and ask those references one question: "what's the gap between quoted and actual delivery?" Factories run late everywhere sometimes; the useful number is how large the gap usually is.
2. Capacity honesty. Ask directly: "If I order 2,000 units in September, does that queue behind anything?" A factory that says "we can always handle it" is telling you they don't track their queue. Ones that describe their busy season, their line constraints, and their cutoff dates for pre-holiday orders are ones that actually plan.
3. Communication latency. Time their responses during the vetting phase, including after they've quoted. A factory that takes 5 days to answer a quote request will take 5 days to tell you production slipped. Communication speed is a proxy for how early you'll learn about problems.
4. Transparency about the factory itself. Will they do a live video walkthrough of the production line? Will they share which processes are in-house versus subcontracted? Refusal on both is a pattern you'll meet again at QC time.

Run every serious candidate through a simple scorecard over the first few orders — quotes vs. actuals, defect rates, response times — so the comparison is data instead of impressions. The structure we use is laid out in the supplier scorecard guide.

Closing the sample-to-production gap

The single most consistent failure in re-sourcing: the sample is beautiful, production is not. Samples are made by the factory's best people, sometimes on different tooling, almost always with more attention than a 2,000-unit run gets. To close the gap:

• Order a pilot run, not just a sample. 50–100 units through the normal production process, inspected against your acceptance criteria. It costs more per unit and tells you more than ten free samples.
• Freeze a golden sample. A signed-off production-grade unit kept by you and one kept at the factory, referenced on every PO. Disputes shrink dramatically when "like the golden sample" is a physical object, not a memory.
• Specify the measurable things. Materials, weights, tolerances, packaging, carton markings, label placement. Anything not written down gets decided by whoever is running the line that day.
• Ask what changed between sample and production. If the pilot run's materials differ from the sample's — different zipper, different foam — you want that disclosed by them, not discovered by you.

Red flags that predict late or bad orders

• Quote in hours for a custom product. Real factories need to check materials and capacity; instant precision is a sales function, not a production one.
• Payment only by untraceable channels, or a big shift in payment details between deposit and balance. Match the payment details to the verified company identity every time.
• Certification claims without documents. Ask for the actual test report PDFs and check the lab and dates.
• The factory resists third-party inspection. A factory confident in its QC welcomes a stranger with a checklist; resistance is data.
• "No MOQ" on a custom product. Genuine custom manufacturing has setup costs; a no-MOQ quote usually means a trading company reselling stock, which changes your quality control and exclusivity picture.
• Everything is included for free (tooling, samples, packaging design). Costs don't disappear; they get recovered later, usually at the worst moment.

None of these are individually disqualifying, but two or more together change how much verification you need before sending money.

Running old and new suppliers in parallel

Don't switch cold. The safe pattern: qualify the new factory with a small order while your incumbent keeps running base volume, then shift the split as the scorecard data comes in — 20% new / 80% old, then rebalance each quarter. This protects your bestsellers while the new relationship proves itself, and it gives you negotiating leverage with both (the terms side of that conversation lives in negotiating with suppliers — this article won't re-tread it).

Worked example, illustrative numbers: you sell 900 dog harnesses a month at a blended 35-day actual lead time. The new factory quotes 25 days. Qualification order of 300 units takes 31 days actual — promising, but not yet 25. You shift 20% of volume (180 units/month) and keep monitoring. Two more orders land at 26 and 27 days. Now you have a measured ~27-day supplier, and you can plan the split and the reorder points from data instead of the quote — and your safety stock requirement drops with every day of verified lead-time improvement.

One tooling note, since lead-time honesty is our home turf: InventoryX takes a manual lead-time number per supplier and builds the weekly buying list from it, so during a parallel-supplier transition you can carry both factories' measured times side by side. It won't discover the new factory for you, obviously — the vetting above is still your job.

Common questions

How do I find a manufacturer for an existing product? Start from constraints, not catalogs: the lead time, volume, and quality floor your current operation requires. Then source through niche directories, trade shows, B2B marketplaces, and referrals from your current factory. Vet for reliability — reference customers at your volume, capacity honesty, response latency — because "can make it" is abundant and "delivers on schedule" is rare.

How do I vet a manufacturer's reliability before a big order? Ask for reference customers at roughly your volume and ask them the gap between quoted and actual delivery. Request a pilot run of 50–100 units through normal production, inspect it against written acceptance criteria, and time how fast the factory communicates during vetting — slow answers in sales predict slow answers when production slips.

Why is production quality worse than the sample? Samples are made by a factory's best people with extra attention; production runs are made at speed. Close the gap with a paid pilot run through the normal process, a frozen golden sample kept by both sides, and written specs for materials, weights, and packaging so nothing is decided by whoever runs the line that day.

Should I switch suppliers all at once? No. Run a parallel transition: qualify the new factory with a small order while the incumbent keeps base volume, shift 20% first, and rebalance quarterly based on measured performance. That protects your bestsellers while the new relationship proves itself and gives you leverage with both suppliers.

Is a lower quote always the better manufacturer? No. A quote you can't verify is a guess, and the cheapest bid often hides the costs in longer actual lead times, higher defect rates, or unfavorable payment terms. Compare factories on landed cost per sellable unit — including defect allowances and the carrying cost of their real lead times — not on the invoice line.

Next step

Pick the product whose supplier worries you most and write its requirement sheet: needed lead time (measured, not quoted), volume profile, and top defects. Then contact three candidate factories with the same sheet and the same questions — comparable answers are the only ones worth comparing.