Sourcing

When to Split One Promo Program Across Two Factories

A mixed promo program looks simple until process ownership is unclear

A distributor is quoting a 12,000-piece event program shipping FOB China to one US fulfillment point in six weeks: 3,000 soft enamel pins, 3,000 woven patches, 3,000 polyester lanyards, and 3,000 zinc alloy keychains. The default move is one PO to one supplier because it reduces onboarding, keeps approvals in one thread, and simplifies carton marking and freight booking.

That approach only works when the quoted supplier controls the actual manufacturing routes. Few factories are equally strong across die-struck metal, die-cast zinc alloy, woven textile, and stitched or printed webbing. A metal factory may run pins and keychains in-house but outsource lanyards and patches. A lanyard factory may do the reverse. The buyer still sees one commercial quote, but sample control, in-line inspection, remake decisions, and production priority are now spread across two or three workshops.

That does not make the one-supplier model wrong. It changes the test. The issue is not whether one factory or two sounds cleaner on paper. The issue is whether one accountable supplier has enough category-level control to protect your schedule, approved specifications, and defect target. On mixed promo programs, the decision usually comes down to process risk, MOQ efficiency, and whether one delayed SKU can block the full shipment.

Quote by manufacturing route, not by shared artwork

A shared event logo does not create a shared production process. Soft enamel pins and zinc alloy keychains require tooling, stamping or die casting, trimming, deburring, polishing, plating, color fill or print, curing, and hardware assembly. Woven patches depend on loom gauge, yarn library, border finishing, cut-edge control, and backing lamination. Polyester lanyards depend on webbing denier and thickness, print method, cutting, folding, stitching or crimping, and breakaway or buckle assembly.

Because the routes differ, the first failure points differ. On a 35 mm soft enamel pin, buyers usually see fine-line collapse below 0.20-0.25 mm, uneven enamel recess fill, plating shade drift, or clutch placement outside tolerance. On a 50 mm zinc alloy keychain, common defects are sharp burrs, jump ring gaps over 0.30 mm, weak split-ring springback, and paint chipping on raised edges after assembly. On a 20 x 900 mm lanyard, the recurring issues are Pantone drift, print repeat misregistration beyond +/-1.5 mm, crooked fold-and-stitch finishing, and hook pull-out below the agreed load. On woven patches, small text, narrow borders, and irregular shapes usually fail first, especially when letter height drops below 2.5-3.0 mm or the edge treatment distorts the finished outline.

That is why mixed programs should be quoted and managed by process family. If a supplier cannot discuss line width, plating thickness, webbing weight, edge method, pull-test level, and AQL by SKU, they are selling commercial coverage rather than controlling production.

Item	Typical 2026 FOB China MOQ / price	What usually fails first	Specs to lock before PO
Soft enamel pin, 30-40 mm	MOQ 300-500 pcs; USD 0.42-0.78 each at 3,000 pcs	Uneven fill, fine-line loss, plating tone variation	Brass or iron base; thickness 1.2-1.5 mm; min line width 0.25 mm; recess depth 0.18-0.25 mm; attachment position tolerance +/-0.8 mm; plating thickness 0.03-0.08 microns for decorative finish; AQL 0.65 critical / 1.0 major / 2.5 minor
Zinc alloy keychain, 45-60 mm	MOQ 300-500 pcs; USD 0.68-1.35 each at 3,000 pcs	Sharp edges, loose jump ring, chipped paint on high points	Zamak 3 or equivalent; thickness 2.5-3.5 mm; split ring wire dia. 1.6-1.8 mm; jump ring gap less than 0.30 mm; edge radius and polish standard approved; paint adhesion tape test; 24-48 hour neutral salt spray if plated
Woven patch, 60-80 mm	MOQ 100-300 pcs; USD 0.34-0.72 each at 3,000 pcs	Text detail loss, border distortion, backing peel	Size tolerance +/-1.5 mm; min readable text height 3.0 mm; woven density approval; merrow, hot cut, or laser edge; backing type; yarn substitution approval; AQL 1.0 / 2.5 / 4.0
Polyester lanyard, 20 x 900 mm	MOQ 100-300 pcs; USD 0.38-0.82 each at 3,000 pcs	Color mismatch, crooked print, hardware pull-out	Webbing thickness 0.8-1.2 mm; width tolerance +/-0.5 mm; print method specified; print alignment tolerance +/-1.5 mm; stitch count or crimp type; breakaway spec; hook or crimp pull test 7-10 kgf minimum

Let the operationally critical SKU drive the sourcing structure

Not every item carries the same business risk. Pins and keychains are usually collectible or promotional. Lanyards are often operational items tied to event registration, badge access, or exhibitor credentials. If lanyards miss the venue date, the program can fail even if the metal items are flawless.

Typical 2026 bulk lead times after sample approval are 10-14 calendar days for standard polyester lanyards, 12-18 days for woven patches, 15-22 days for soft enamel pins, and 18-25 days for zinc alloy keychains when tooling, plating, and assembly are straightforward. Add 3-5 days for a pre-production sample, 2-4 days for export packing and delivery to port or consolidator, and 3-7 extra days in March-May and August-October when plating lines and looms are busier.

In a one-supplier model, the fastest line often waits for the slowest. A remake on metal due to enamel color rejection, a failed plating adhesion test, or a tooling correction on keychains can hold the whole booking because the supplier wants one final ship date. Ready lanyards then sit packed while the slowest SKU catches up. For a fixed event date, that is unnecessary schedule concentration.

A simple rule works well. If one SKU is operationally critical and another is mainly decorative, do not let the decorative SKU control the vessel booking. If synchronized arrival is not essential, split sourcing usually reduces schedule risk.

Build the cost model beyond the unit price

One-stop quotes often look efficient because subcontract margin is buried inside the line item. On mixed promo programs in 2026, that hidden premium is commonly 8-20 percent on categories outside the supplier's core process. On 3,000 lanyards quoted by a metal-focused supplier, the markup is often USD 0.08-0.18 per piece. On 3,000 woven patches, it is often USD 0.05-0.14 per piece. Across the textile portion alone, that can add USD 390-960 without improving quality or lead time.

Two factories do create overhead. You may pay two sample courier charges, two document handling fees, and two local China deliveries to a consolidator. If the goods move on separate destination entries, there may also be an extra customs brokerage fee. In practice, on 3,000-piece-per-SKU programs, that overhead is often lower than the hidden subcontract margin or the cost of missing a fixed launch.

Cost element on a 12,000-piece mixed program	One supplier model	Two factory model
Outsourced-category margin	Often hidden; typically USD 300-900 across the program	Usually avoided or visible in direct specialist quotes
Sample and approval cost	Usually lower by USD 50-150	Usually higher by USD 50-150
Local China consolidation cost	Lower if one shipper handles all cartons	Often USD 80-220 extra if two factories deliver to one consolidator
Risk of one SKU blocking full shipment	High	Lower
Rework containment	Harder when root cause sits with a subcontractor	Cleaner by category, with direct accountability
Best FOB outcome at 3,000 pcs per SKU	Often competitive on core items only	Usually stronger on mixed metal plus textile programs

There is still a scale threshold. When each SKU is under 500 pieces, setup charges, admin time, and sample cost can outweigh specialist savings. Once each SKU is above roughly 2,000 pieces and the order spans both metal and textile processes, specialist sourcing usually performs better on cost transparency and defect containment.

Pressure-test the one-supplier option for actual process control

The weak question is, 'Can you make this?' Most suppliers will say yes. The useful questions are operational: what is produced in-house, what is subcontracted, who owns pre-production approval, who performs in-line inspection, who signs final AQL, and what corrective action happens before packing if defects exceed target.

For metal items, ask for evidence from the actual route: die-struck or cast blanks, polishing standard, plating line capability, enamel fill method, curing temperature range, and 100 percent attachment inspection. For lanyards, ask for webbing thickness or gsm, print method by artwork type, stitch pattern or crimp construction, breakaway specification, and pull-test records. For woven patches, ask which loom gauge is used, how yarn substitutions are approved, what edge finish suits the shape, and how warped or undersized pieces are segregated before packing.

A supplier that really controls the category should give realistic MOQ tiers, sample lead times, bulk lead times, packaging options, carton limits, and inspection standards without vague language. If every answer is 'no problem' but nobody will commit to numbers, tolerances, or ownership points, you are dealing with a commercial integrator, not a process owner.

• Require every supplier to declare each SKU as in-house or subcontracted before quote approval
• Ask for separate lead times for artwork proof, pre-production sample, bulk production, and export packing by SKU
• Lock category-specific AQL targets instead of using one generic quality clause
• Approve one sealed sample or signed reference image for each process family: metal, woven textile, and printed webbing
• Confirm packing method, inner-pack quantity, carton size and weight limits, and whether mixed-SKU master cartons are allowed
• Set an escalation trigger if any SKU slips more than 3 calendar days against the approved plan

Use one master spec pack even when production is split

Splitting factories does not create inconsistency by itself. Weak specifications create inconsistency. A proper mixed-program pack should include one master brand control sheet and separate technical sheets by SKU. The master sheet should cover Pantone references, approved logo lockups, copy, finish hierarchy, barcode or insert rules if needed, carton marking format, and shipping mark layout. Each item sheet should then cover the limits that matter to that process.

For example, Pantone 186 C may be the campaign red, but approval cannot stop at the Pantone callout. On a soft enamel pin, that red can read darker after baking and appear deeper next to black nickel. On a woven patch, the nearest stock yarn may appear slightly duller because yarn libraries are discrete rather than continuous. On a dye-sublimated lanyard, the same red may print brighter on white polyester. Those are substrate effects, not automatic defects.

The practical fix is to define visual tolerance by item role. If the pin is the hero collectible, let it anchor the appearance standard. The patch and lanyard then need to remain visually aligned under standard indoor lighting such as D65-equivalent office lighting, not match perfectly across different materials. That reduces unnecessary remake arguments over cross-substrate differences no competent factory can eliminate completely.

The pack should also include measurable packaging rules. Typical examples are 50 pcs per inner polybag for lanyards, 100 pcs per inner for patches, individual polybags for pins only when plating or epoxy protection requires it, outer carton gross weight capped at 12-15 kg, and no mixed hardware finishes in the same inner pack. Those details matter when two factories feed one fulfillment point or one China consolidator.

Choose the split point by accountability, not by product type alone

There are three common models for this type of order. Model one is a single accountable supplier managing some outsourced categories. It works when quantities are modest and the delivery schedule has at least 7-10 buffer days. Model two is two specialist factories with buyer-managed coordination. It works when internal sourcing bandwidth is strong and process control matters more than admin simplicity. Model three is a lead supplier plus one direct specialist, which is often the most practical middle ground.

For the 12,000-piece event scenario, model three is usually the strongest structure. Put lanyards with a textile specialist because they are date-critical, comparatively fast to produce, and sensitive to print and assembly control. Put pins and keychains with a metal specialist because tooling, plating consistency, edge finish, and attachment integrity matter more there. Place woven patches with whichever supplier proves better loom detail, backing control, and edge finishing through sample evidence rather than artwork similarity.

This split also fits normal factory economics. Metal suppliers usually quote better when pins and keychains share plating finish, backing card format, or polybag method. Textile suppliers usually quote better when lanyards and patches share color approval timing and export packing flow. Bundling everything under one supplier only makes sense when that supplier can prove category-by-category control, provide competitive direct pricing on non-core items, and still meet the schedule without using the event date as the production buffer.

Run the RFQ in a way that exposes risk early

Start with one itemized RFQ pack sent to at least one metal specialist, one textile specialist, and any one-stop supplier under consideration. Break out tooling, sample charge, unit price, packing, carton specification, production days, and FOB terms for each SKU. Require every supplier to identify which lines are in-house and which are subcontracted. Without that disclosure, you cannot compare either price or risk honestly.

Then score each response on four factors: schedule confidence, process control, transparency, and total program cost. A simple weighted model works well for event programs: 35 percent schedule confidence, 30 percent process control, 20 percent total cost, and 15 percent transparency. That weighting reflects reality. A late operational SKU is usually more expensive than a small unit-price saving.

Before PO release, freeze the control points that actually protect the order: one master brand sheet, one approved sample per process family, one packaging matrix, one final AQL by SKU, and one escalation rule. A practical rule is this: if any SKU is forecast to slip more than 3 calendar days against plan, immediately review split shipment, partial air upgrade, or revised carton-level consolidation.

The sourcing decision is straightforward once the routes are visible. Split the program when process families differ, quantities are meaningful, and one SKU carries higher schedule risk than the others. Keep it with one supplier only when that supplier can prove operational control across categories and the schedule has enough buffer to absorb a miss. On mixed promo orders, the cheapest-looking quote is often not the lowest-risk structure.

Have a project? Send your artwork and target quantity and we’ll reply with a detailed quotation within 12 working hours.