The Decision Is Made: Now What
Move a golf bag program between suppliers as a ninety-day project, not a purchase-order swap: transfer tooling and IP deliberately, hand over the knowledge a tech pack cannot carry, run the new house in parallel, validate with a first order against the golden sample, and leave the old relationship intact — most transition failures are project failures, not supplier failures.
The statistic the trade lives with, stated as a pattern rather than a number: half of supplier transitions disappoint, and almost none disappoint because the new supplier could not technically make the bag. They disappoint because the tooling arrived late, the spec knowledge lived in one engineer's head at the old house, the first order shipped against optimism instead of the golden sample, the old supplier stopped trying two weeks before the last container, or the customer-facing channel noticed the seam change nobody told it about. Every one of those is a project-management failure wearing a sourcing costume — which is the good news, because project failures are preventable with a playbook.
The playbook's shape, before the details: a transition is a product launch and a product retirement happening simultaneously, with the buyer's brand riding both. The launch side (the new house's first order, validated like any launch — sampled, audited, AQL-checked) and the retirement side (the old house's last order, run with MORE care than the average one, not less) are equally important and usually managed with opposite emotions. The discipline that follows treats both sides as the same project, on one calendar, with one owner — because the transitions that fail usually had two owners who each assumed the other owned the seam.
The Transition Project, Scoped
The scope that fits one page and governs ninety days: the program inventory (which SKUs move — the entire line, the problem styles, or the new-season styles only; the honest answer is usually a phased list, because moving the flagship and the slow movers in the same quarter doubles the project for no additional risk reduction), the calendar (anchored on the season's immovable dates — the buying calendar's set dates work backward through first-article approval, pilot production, and the new house's own learning curve, and the project that ignores the calendar discovers it in November), and the two-house budget (the parallel-run costs, the duplicate tooling where transfer is impossible, the expedite fees the gap will charge — the honest line items that make the transition's true cost visible before it is spent).
The scoping question most programs skip, and pay for: what does ‘done’ mean? The transition is complete not when the new house ships something, but when it ships the season's full order, to spec, on its own calendar, without the old house's involvement — and the intermediate gates are worth writing down (pilot samples approved, first production order passed, second order on standard terms, old house's obligations discharged). A transition scoped this way has a finish line; a transition scoped as ‘move the bags to the new factory’ has a feeling, and feelings do not close projects.
What Actually Moves: Tooling and IP
The physical and legal inventory, taken before anything ships: the tooling the program owns (cutting dies, molds for badges and base components, embossing plates — the items the ownership terms were written about, and the reason those terms existed), the digital assets (the graded patterns, the marker files, the color standards — the artifacts that reproduce the product without a conversation), and the paper (the signed golden samples, the tech pack in its current revision, the test reports). The transfer request is a formal letter under the agreement's return clauses, with a manifest, a date, and a condition log — because tooling that arrives dinged or files that arrive corrupted are the transition's most common week-three surprise.
The two honest complications: first, some tooling never moves (worn dies and battered molds often cost more to transfer and revalidate than to remake — the remake is also the new house's calibration exercise, which is worth more than the salvage; the decision is per-item, priced both ways). Second, the IP conversation and the exit conversation must not collide: the house being exited holds patterns and samples it made, and the agreement's surviving obligations (confidentiality, the return or destruction of brand property) are exercised precisely, in writing, at the transition's start rather than its end — the awkward conversation early is a paragraph, and the same conversation late is a dispute. The new house signs its own protection terms before the first file transfers, not after the first sample impresses.
The Knowledge Transfer Paper Cannot Do
The tech pack carries the specification; it cannot carry the program's muscle memory — the hundred small preferences that made the old house's version of the bag feel the way it felt: the strap's exact fold at the anchor (the pattern shows the seam, not the touch), the logo thread tension that reads correctly in photographs (the spec sheet says ‘embroidered, 7,500 stitches’, not ‘the density that photographs like brand quality’), the base-plate bonding sequence the old line evolved over three seasons. This knowledge transfers three ways, and all three are on the project plan: the annotated golden sample (the physical reference, marked where the feel lives — the one artifact that argues without vocabulary), the video library (the hour of phone footage worth more than a binder — the old house's best operator running the tricky operations, narrated; most houses will film this if asked plainly, because it is a compliment), and the engineering call (the new house's sample master talking to the old house's sample master, one arranged call, questions prepared — the cheapest consulting either house will ever provide).
The knowledge transfer's ugly twin, equally on the plan: the knowledge that should NOT transfer. The old house's workarounds (the tolerance that drifted, the shortcut that a season of passed inspections quietly licensed) are also craft knowledge, and a transition done naively copies both. The clean break is deliberate: the new house builds to the spec and the golden sample, not to the old house's habits — and the first-article comparison (new sample against golden, dimension by dimension, the audit discipline applied to the artifacts) is where the program learns which differences are the new house's learning curve and which are the old house's quiet drift it is accidentally escaping.
The Parallel-Run Economics
The bridge that most programs run and none budgets honestly: the period when both houses hold orders, and the transition's real cost lives there. The parallel period's arithmetic, in the open: the duplicated fixed costs (two sets of samples, two minimums met, two first-article cycles — the price of buying certainty), the deliberate split (the pilot styles at the new house while the volume styles finish at the old — the risk staged instead of bet, and the premium the staging charges is the transition's insurance premium, priced honestly at 5-15% of a quarter's program cost), and the sunset date (the parallel period has an end written into the plan, because the house being exited behaves differently once it believes it has already ended — the sunset keeps its effort alive through the last container it will ever run well).
The parallel-run's quieter value, worth more than the risk math: the comparison data. The program that runs both houses through the same season holds, for one quarter, the controlled experiment the trade never otherwise gets — the same specs, the same test methods, two factories' interpretations, and a defect-and-delivery record that makes the exit decision's wisdom measurable. The programs that skip the parallel run to save the premium usually pay it later at the claim line, minus the data.
The First Order as the Final Exam
The new house's first production order is not an order; it is the transition's final exam, and it is graded like one: the pre-production gate (the pilot run at production speed, on production lines, with production workers — not the sample room's best hands; the pieces that survive contact with the real line are the data), the inspection regime turned to maximum (in-line checks plus a tightened final AQL — the new relationship's first impression being worth the extra inspection cost, and the third-party inspection fee being the cheapest grade of insurance the transition buys), and the comparison discipline (the first order's pieces against the golden sample with calipers, not vibes — the receiving inspection at the dock doubling as the transition's report card).
The pass/fail logic, agreed before the exam starts: the first order that passes within tolerance graduates the new house to standard terms (the standard calendar, the standard inspection regime — the graduation being the point, because the new house that stays on training wheels never learns to walk); the first order that fails gets the remediation conversation with the same evidence discipline the old house ever received (findings, photographs, cause, countermeasure — the crisis discipline at transition scale), and the second first-order. What the first order never gets is the benefit of the doubt — the transition that grades softly has merely moved its future claims to a new address.
The Exit Etiquette With the Old House
The house being exited is owed an exit, and the etiquette is strategy rather than politeness: the last order runs at full discipline (the temptation — and it is real — is to coast the final order; the counterargument is that this order's customers are the same customers the new house will serve, and the brand that lets its last old-house order leave carelessly is mailing its own reputation to a stranger's address), the obligations are discharged in writing (the final payment on time, the tooling returned or the terms for its disposition executed, the surviving confidentiality terms acknowledged — the agreement's endgame played by the book, because the exit clause was written for exactly this Tuesday), and the conversation is honest without being final (houses explain the transition as capacity strategy, cost architecture, or program evolution — any true sentence that lets both teams keep their dignity, because the industry is small and the next chapter is long).
The strategic reason the etiquette pays: the exited house remains a capability in the market. The program that exits well keeps a warm second source (the house that was treated well in the exit answers the emergency call eighteen months later — the dual-source option that costs nothing but the exit's manner), while the program that exits badly (the tooling held hostage over a disputed invoice, the team told through a forwarded email) converts a former partner into a permanent negative reference in a trade where every buyer's shortlist is three calls long. The exit etiquette is the cheapest reputation insurance a program ever buys — and the only line item on the transition budget that consistently returns its cost.
The Supply Gap Risk
The transition's most dangerous failure mode is not the new house's quality — it is the gap: the weeks or months when the old house has stopped and the new house has not started, and the channel's shelves quietly empty. The gap is governed by the program's demand math: the safety stock that bridges the transition is sized against the gap's honest duration (the new house's ramp curve — its first order's calendar padded for the learning it is doing — plus the inspection and freight legs, plus the two weeks everything takes longer than planned), and it is built deliberately (the bridge inventory ordered from the old house as a named final production run, not discovered as whatever happened to be on the shelf when the calendar turned).
The gap's secondary risk, which the bridge inventory creates: the aged bridge. The stock that bridges a long gap is the stock that sits through a season change (the storage clocks running on the program's buffer), and the bridge that is over-built converts the transition's insurance into next year's clearance. The honest sizing rule: bridge to the new house's second order, not its fifth — the gap risk is real, but so is the aged-inventory risk, and the transition plan that prices both is the plan that survives contact with the calendar.
What the Customer Should Know, and When
The channel-facing side of the transition, handled with the same deliberateness as the factory-facing side: most programs tell their channel nothing (the seam change nobody mentions, discovered by a customer comparing two seasons' bags — the perception of quality drift created by a change that was managed perfectly on the factory side), and a few programs tell their channel everything (the announcement that invites the channel to worry about stability it was never going to notice). The working middle is the change-note discipline: the channel gets told what changed when the change is customer-visible (a hardware switch, a lining revision — the honest one-paragraph note that says ‘the 2027 model carries X instead of Y, for these reasons’), and gets told nothing when the change is genuinely invisible (the same bag, built across town — which is the transition's silent majority).
The timing rule that keeps the notes honest: the channel learns at the point the change becomes purchasable, not before (the pre-announcement that invites questions nobody can answer yet) and not after (the customer service team discovering the change from a return). The transition plan's communication gate sits between the first-article approval and the first container's shipping — the moment the change is real, proven, and safe to describe — and the note that ships then is drafted once, approved by both the product and the commercial sides, and becomes the season's single source of truth. The channels forgive changes they are told about; they remember the ones they discover.
The Double-Source Steady State
The end-state worth designing toward, because some transitions are better as permanent architectures: the program that needs resilience (the single-source fragility the exit just exposed) or flexibility (peak seasons that no one house can absorb) may deliberately stop at two houses — the primary carrying the volume and the flagship, the secondary carrying the overflow and the peak tranches, both houses kept warm by the same tech pack and an annual cross-order. The double-source steady state has real costs (two relationships to maintain, two quality files, the split's fixed minimums) and real rules (the golden sample discipline enforced identically at both — the consistency question is the double source's central question), and the programs that run it deliberately treat it as an insurance product with a known premium.
The steady state's honest test, worth applying before committing: does the second house ever ship? The double source that never produces is a paper exercise (a relationship maintained in quotes and courtesies, discovering its unreality at the exact moment it is needed — the peak season that reveals the secondary house has moved, retooled, or forgotten the program); the double source that ships a real tranche annually (the pre-book's deliberate split, the overflow that is genuinely routed) is a live capability. The annual cross-order is the double source's heartbeat — the single order per year that keeps the second house's hands on the program, and the cheapest resilience premium the market sells.
A Transition Run Well, Worked
The playbook exercised on a composite program — a mid-market stand-bag line moving after two seasons of delivery drift: the scoping (the entire six-SKU line, phased — the three core styles at the new house first, the three slow movers following a season later; the calendar anchored on the February set date, working backward through the December pilot, the October tooling transfer, the September house-selection close), the transfer (the cutting dies moved with a manifest and photographed condition log; the badge molds remade instead of moved — the remake doubling as the new house's calibration; the annotated golden sample plus four hours of operation footage from the old house, narrated by its line lead as a courtesy priced at a dinner), and the parallel quarter (the bridge inventory built as a named final order — seven weeks of demand from the old house, stored to the storage band while the new house learned).
The exam and the landing: the pilot ran in December (eleven pieces on the production line, three failed the strap-anchor pull test — the remediation conversation held with photographs, the fix made, the re-run passed), the first production order shipped in February against a tightened AQL and a third-party inspection (passed at 98.4%, four minor defects documented and dispositioned, the dock check confirming the golden-sample match), and the exit ran by the book (the last payment on time, the tooling returned on the manifest, the honest capacity-strategy conversation held — and the old house answering the emergency call eleven months later when a competitor's collapse freed up a tranche the new house could not absorb). The program's verdict, one year on: the transition cost 9% of a quarter's program cost all-in, the delivery drift was gone, and the second source was warm — the ninety days, run as a project, had done what the decision alone never could.
The Relationships Left Standing
The closing audit the mature programs run, because a transition is also a relationship event and relationships compound: the old house, treated well through the exit, remains a warm capability (the annual courtesy call, the shared trade-show coffee, the industry reference that answers honestly — the network effect of an exit run with etiquette); the new house, onboarded through a project rather than a leap, starts the relationship at year-two trust (it has seen the program's discipline, answered the hard questions, and passed a real exam — the onboarding that principles of good onboarding teach from the channel side, applied to the factory side); and the buyer's own team has a playbook, written and proven — the next transition starting at page one of a document instead of at zero.
And the quiet principle the whole playbook compresses into: sourcing is a portfolio of relationships, not a series of transactions. The program that changes houses well — project-managed, knowledge-transferred, parallel-bridged, exited with dignity — is building the only supply-chain asset that survives every other change: a market's worth of houses that would work with it again. That asset is what the ninety days are actually for; the bags are merely what they produce along the way.
Frequently Asked Questions
How do I switch golf bag suppliers safely?
Run the change as a ninety-day project: scope it on one page (which SKUs, which calendar dates, what done means), transfer tooling and IP with a manifest and the agreement’s exit clauses, hand over the knowledge a tech pack cannot carry (annotated golden sample, operation footage, one engineering call), bridge the gap with a named final order from the old house, and validate the new house with a pilot run and a tightened-AQL first order against the golden sample.
What should transfer to a new golf bag factory?
The program’s owned tooling with a photographed condition manifest, the current-revision tech pack and graded patterns, the signed golden samples, and the color standards. Decide per item whether worn tooling moves or is remade — a remake doubles as the new house’s calibration. The new house signs its own protection terms before the first file transfers.
How long does a supplier transition take?
Plan ninety days from decision to first clean container for a program moving within a familiar category: tooling and file transfer in the first month, pilot samples mid-project, the first production order validated against a tightened AQL before the standard calendar resumes. Larger line moves phase over two seasons — the calendar’s immovable set dates are the anchor everything works backward from.
Should I run both suppliers in parallel during a switch?
Usually yes — the parallel period is the transition’s insurance and its only controlled experiment. Budget it honestly at 5-15% of a quarter’s program cost (duplicate samples, two minimums, two first-article cycles), stage the risk by piloting new-house styles while volume finishes at the old, and write the sunset date into the plan so the exiting house keeps its effort through the last container.
What is a bridge order in a supplier transition?
The named final production run from the exiting house that covers the gap between the old house stopping and the new house shipping — sized against the new house’s honest ramp curve plus inspection and freight, plus two weeks of reality. Bridge to the second order, not the fifth: the gap risk is real, but over-built bridges age into next year’s clearance.
How do I transfer production knowledge a tech pack cannot carry?
Three artifacts, all on the project plan: the annotated golden sample (the physical reference marked where the feel lives), the operation-footage library (the old house’s best operator running the tricky operations, narrated), and one arranged engineering call between the two sample masters. Guard the other direction too — the new house builds to spec and golden sample, not to the old house’s drifted habits.
How should I treat the supplier I am leaving?
With full-discipline exit etiquette, as strategy: run the last order harder than average, discharge every obligation in writing and on time, and explain the change honestly without finality. The exited house remains a warm second source and a market reference — the exit that costs nothing is the cheapest reputation insurance a program ever buys.
When should customers be told about a factory change?
At the moment the change becomes purchasable — after first-article approval, before the first container ships — and only if the change is customer-visible. The change-note discipline: one honest paragraph on what changed and why (a hardware switch, a lining revision), approved by product and commercial sides together. Channels forgive changes they are told about and remember the ones they discover.
Is dual sourcing worth keeping after a transition?
If the program needs resilience or peak flexibility, yes — deliberately, as a priced insurance product: the primary carries volume and flagship, the secondary carries overflow and peak tranches, both held to the same golden sample. The test of a live double source is whether the second house ever ships — one real tranche annually is the heartbeat that keeps the capability warm.
What can go wrong in a supplier transition?
The pattern is project failures, not capability failures: tooling arriving late or damaged, spec knowledge trapped in one engineer’s head, the first order graded softly, the exiting house coasting its final run, and the channel discovering a visible change on its own. Every one is prevented by the playbook’s gates — manifest, knowledge artifacts, tightened first-order AQL, sunset date, change-note discipline.
Do I need a new manufacturing agreement with the new supplier?
Yes — signed before the first file transfers, not after the first sample impresses. The agreement’s ownership, confidentiality and exit clauses are precisely what makes this transition reversible and clean; the new house’s protection terms and the tooling provisions should reflect everything the old relationship taught you about how agreements behave on a bad day.
How much does a supplier transition cost?
Budget the honest line items: duplicated sampling and minimums, the parallel-run premium (5-15% of a quarter’s program cost where staged), the bridge inventory’s carrying cost, third-party inspection on the first order, and expedite fees. A composite worked transition landed at 9% of a quarter’s program cost all-in — and priced against one missed set date or one season of delivery drift, it is the cheaper side of the ledger.