Bring Outsourced Production Back In-House: A Capacity-and-Cost Test

Bringing outsourced production back in-house requires two tests: can the complete qualified route deliver the job, and will avoidable supplier spending exceed incremental internal cost? Idle time on one machine answers neither question alone. Test the capacity and cost of one stable job family, then run a bounded trial with a supplier fallback. Keep quality approval, delivery dates and existing customer commitments in the comparison.
Release rule: approve the next lot from accepted output and actual cost, with procurement confirming which supplier charges can disappear.
| Test | Evidence | Decision owner |
|---|---|---|
| Complete-route capacity | Job, setup, people and inspection hours | Operations and planning |
| Quality readiness | Qualified tooling, route and acceptance | Quality |
| Avoidable spending | Supplier terms and confirmed releases | Procurement and finance |
| Trial continuity | Bounded lot and fallback commitment | Planning and procurement |
Choose one subcontracted job and an equivalent delivery promise
Start with one stable job family whose route, specification and supplier commitments are known. A broad goal to reduce outsourcing is too loose for a make-or-buy decision.
Name the parts, revision, lot size, accepted quantity and delivery date. Record which operation is outsourced and which steps already happen internally. Compare alternatives against the same customer requirement, rather than comparing the supplier's finished accepted parts with an internal machine's unfinished output.
Choose work that can be observed through a bounded trial. A repeated job with familiar tooling and inspection may be easier to qualify than a new, volatile product. The best candidate is not necessarily the largest supplier invoice. It is the job with a credible internal route and spending that can actually be removed.
The NIST Manufacturing Extension Partnership's sourcing guidance emphasizes total cost and continuity of supply. Those principles also matter when work returns in-house. The supplier's flexibility and fallback capacity can have value while the internal route is being proved.
In Guidewheel's Seacast case study, outsourcing could cost around $40,000 per month. The story describes improved setup-time visibility and plans to reduce reliance on outsourced work. It does not report that $40,000 as realized monthly savings. Use that story to frame the operating question: which specific work could be delivered internally, and what spending would disappear if it were?
Have procurement document cancellation windows, minimum purchases and supplier-held inventory before calculating savings. Existing commitments remain real costs even when the production plan changes.
Check capacity across the complete qualified route
Internal capacity means available qualified hours across every required step. Spare time on the first machine is not a delivery commitment.
Map setup, production, finishing, inspection and packing. Add the people, fixtures, tools and materials that each step requires. Check whether the internal route meets customer approvals and quality requirements before allocating it production volume. A fast machine cannot compensate for an unqualified process or unavailable inspection capacity.
This illustrative trial lot requires 1,000 accepted units on the same promised date as the supplier alternative. The assumed route requirements include an allowance for rejects and their extra work:
| Route step | Processing hours | Setup hours | Total required | Available after existing commitments |
|---|---|---|---|---|
| Primary operation | 40 h | 6 h | 46 h | 50 h |
| Finishing | 20 h | 4 h | 24 h | 28 h |
| Inspection | 10 h | 2 h | 12 h | 16 h |
All three steps fit the illustrative schedule, with qualified people and tooling assumed available. Packing is included in finishing for this example. Match shared labor and tooling to the actual windows, and check that the steps can finish in sequence by the due date. Existing customer jobs remain protected. These available hours are net of their commitments, not total machine calendar hours.
Sensitivity changes the answer quickly. If finishing has only 22 available hours, the lot has a 2-hour route shortfall even though the primary machine still fits. Planning must move permitted work, secure additional qualified hours or keep the supplier lot. It should not call the missing time a minor detail after approving insourcing.
Guidewheel's production-capacity guide introduces the wider capacity question. For this decision, use observed job and setup records, qualified output and route availability together. A current signal can show operation; it cannot alone certify a part or prove that every required route step is available.
Compare avoidable supplier spend with incremental internal cost
Compare the spending that changes between alternatives. Keep existing cost allocations separate so the worksheet can answer whether the company will spend less.
The illustrative supplier invoice is $12,000 for 1,000 accepted units. Procurement confirms that $1,200 remains payable under an existing commitment. Avoidable supplier charges are therefore $10,800. An additional $400 of supplier-route freight can be avoided, bringing the comparison's avoidable external spending to $11,200.
The internal alternative assumes the following incremental costs to deliver the same accepted lot:
| Internal cost | Illustrative amount | Inclusion basis |
|---|---|---|
| Material, including assumed trial loss | $5,000 | Material required by the internal route |
| Additional paid labor | $1,600 | Spending caused by this lot |
| Energy | $300 | Incremental route consumption cost |
| Consumables | $400 | Additional use |
| Inspection | $500 | Incremental acceptance work |
| Disposal | $100 | Incremental loss handling |
| Internal-route freight | $200 | Required delivery spending |
| Recurring total | $8,100 | Same accepted quantity and date |
The recurring advantage is $11,200 − $8,100 = $3,100 per comparable lot. Assume $1,800 of one-time tooling and qualification expense for the initial transition. The first lot's advantage then falls to $1,300. These are illustrative inputs, not supplier quotes, benchmarks or a customer result.
Do not automatically add existing salaried labor, depreciation or facility rent as incremental spending. Finance may need those costs for product costing, but their allocation does not prove they change because this job moved. Equally, do not treat scheduled employee time as free when the work requires extra pay or displaces another committed job.
Compare complete alternatives. If both routes use company-supplied material, remove the common material spending from both sides or include it consistently. Do not compare a conversion-only supplier price with a full internal cost without adjusting the boundary.
Run a limited production trial with supplier fallback
Trial the qualified route before releasing more volume. Preserve enough supplier capacity to recover the delivery plan if the trial fails.
Operations, quality and procurement should agree the trial lot and acceptance conditions before production starts. Specify the accepted quantity, due date, inspection release, allowed extra labor and cost-record requirements. The trial should also identify who can stop it and how the supplier fallback will be activated.
For the illustrative lot, the acceptance record calls for 1,000 quality-approved units, the agreed delivery date and actual incremental costs within the reviewed comparison. The route-hour table is an input to that decision, not evidence that the trial has already passed. Only completed production and quality records can establish its result.
Keep the fallback practical. Confirm what the supplier can still deliver, by when, at what agreed charge and after what notice. Record whether reservation or expedited freight creates a real extra cost. Do not assume an existing supplier can accept work immediately after its orders were canceled.
| Trial record | What the reviewer should see |
|---|---|
| Quantity and disposition | Accepted, rejected and remaining units |
| Complete-route hours | Setup, processing, inspection and waiting |
| Delivery | Actual readiness and shipment against promise |
| Internal spending | Material, paid labor, energy, inspection and logistics |
| Supplier exposure | Retained commitments and fallback expense |
| Release decision | Increase, repeat, modify or return work |
Capture learning rather than declaring success from one machine's runtime. If quality rejects rise, record their material and processing expense once and revise the accepted-output requirement. If upstream trial work creates a finishing queue, use that evidence to resize the next lot. Protecting continuity may justify retaining outsourcing even when the internal recurring comparison looks attractive.
Make the release decision from accepted output and actual cost
Release the next job only when the trial supports the same delivery promise and the cost advantage survives realistic changes.
The illustrative first-lot advantage is $1,300. If the $1,600 incremental labor input rises by 20%, an extra $320 reduces that advantage to $980. If the lot requires additional supplier fallback, tooling or expediting, those charges belong in the revised comparison. A favorable baseline should not be protected by hiding the trial's unexpected work.
For recurring approval, compare the observed accepted yield, complete-route hours and internal spending with the released supplier charges. Retain the $1,200 external commitment in the company-level picture, even though it was excluded from avoided spending. The supplier's gross invoice and the truly avoidable amount should remain visible together.
Use three possible decisions. Increase a bounded volume when quality, capacity and spending all pass. Repeat or modify the trial when a specific uncertainty can be resolved. Keep the supplier route when the internal option cannot meet its qualified-output promise or its spending advantage disappears. A rejected insourcing case is still useful capacity evidence.
Guidewheel's practical operating platform helps create the machine and operating record behind this review. Its AI-powered software and manufacturing experts can help teams identify the next capacity opportunity. Optional system context can connect the record to jobs and other workflows, with the required integration scope confirmed for the application. Procurement, planning, quality and finance retain responsibility for supplier terms, product qualification and release decisions.
Keep the model available after the transition. Product mix, supplier terms, wage costs and route constraints can change. Revisit the comparison when those inputs change or a delivery exception reveals that the original assumption no longer holds. Insourcing is a controlled operating decision, not a permanent label attached to every job once a machine has spare hours.
Make the next scoped decision
Bring one supplier job, its contract exposure and the complete internal route to the review. Agree what the trial must deliver and which supplier commitment protects continuity. To explore how operating visibility and expert support can help establish the capacity evidence, discuss your operation with Guidewheel. Release the next bounded lot from the observed result, rather than translating a supplier invoice into a promised saving.
Frequently asked questions
Are idle machine hours enough to justify insourcing?
No. The full route also needs qualified tooling, people, material, finishing and inspection capacity. Protect existing commitments and verify accepted output within the required delivery window before releasing internal volume.
Which supplier costs count as avoided spending?
Count charges procurement confirms can be removed within the comparison period. Retained minimum purchases, noncancelable commitments and fallback reservations remain costs. Avoidable freight can enter separately when its cancellation is supported.
Should existing overhead be added to the internal alternative?
Include existing allocations when finance needs full product costing, but separate them from spending that changes because the job moves. Incremental paid labor, material, energy and transition costs answer the make-or-buy cash question.
How should supplier fallback work during the trial?
Agree the reserved quantity, notice deadline, delivery window and cost before the trial. Planning should know when to activate fallback, and procurement should confirm the supplier can still meet the commitment.
Did Seacast report $40,000 per month in realized savings?
No. Guidewheel describes outsourcing that could cost around $40,000 per month and plans to reduce reliance on it. The story does not establish that amount as completed monthly savings.