Make or Buy · Outsourcing Decision
Make-or-Buy Decision
What is Make or Buy?
- Four axes: cost, capability, strategic importance, risk
- Keep core-competence areas in-house
- Mixed, phased approaches are more realistic than an either-or choice
Applying Make or Buy properly changes the questions asked in the meeting. The focus shifts from “What’s the unit price?” to “Do we want to own this capability?”, and the decision moves from the purchasing manager’s desk up to the management team. That is because bringing a function back years after outsourcing it often costs more than was saved at the start. So this judgment should be made not against this year’s budget, but against what your organization will still have three to five years from now.
In practice, the answer usually lies somewhere between the two extremes. Common approaches include a split, where design and standards stay in-house and only execution is outsourced; dual sourcing, where the same work is divided between two suppliers; and parallel operation, where 20–30% of volume stays in-house so you see cost and difficulty firsthand. Hybrids between full outsourcing and full insourcing are closer to the norm. You can also split by time. Outsource for the first 1–2 years to validate demand, then reopen the insourcing question once annual volume passes break-even.
Certain items are routinely missing from the comparison. Outsourcing carries management effort on top of the unit price. Writing specifications, inspecting deliverables, settling accounts, and coordinating with the vendor add up to 0.2–0.5 FTE of a staff member on an ongoing basis. If that is left off the cost sheet, outsourcing looks cheaper than it really is. Conversely, in-house work carries, beyond labor costs, a 6–12 month non-productive period for hiring and ramp-up, plus the opportunity cost of those people being unable to do other work. The numbers only mean something once the hidden costs on both sides are in the same table.
The most common failure is “we outsourced it because it was cheap” turning, a few years later, into “we have no one left who can write the requirements.” Buying well is a capability too. If no one in-house can define specifications and judge the quality of deliverables, you cannot verify whether a supplier’s quote and schedule are reasonable, and you lose negotiation leverage at renewal. The rule is to keep the ability to judge, even for non-core work. The opposite mistake is just as common. If you take on everything that looks important, you only add underused fixed costs, while the people who should be focused on the core get tied up in routine tasks.
John KimInstructor John Kim, author of “B2B Marketing Design” and CEO of Nanum Consulting
Jinho DanAccounting instructor who makes financial statements, costing, and tax easy to understand| Criteria | Make (In-house) | Outsource (Buy) |
|---|---|---|
| Capability | Know-how accumulates in the organization | Little internal learning builds up |
| Cost | Fixed cost, expensive at low volume | Variable cost, low upfront burden |
| Speed | Takes time to build | Ready to use right away |
| Control | Direct control over quality and schedule | Depends on contracts and SLAs |
| Good fit | Core competency areas | Non-core and variable-demand areas |
Transaction cost economics (Ronald Coase, Oliver Williamson) supplied the theoretical basis by explaining 'why firms exist.' In practice, it is a basic framework for procurement and strategy decisions.
Consider a hypothetical case. Company A, an industrial parts manufacturer with 540 employees, had outsourced all of its new-hire and experienced-hire onboarding training to external providers. Annual outsourcing fees had climbed to KRW 170 (≈ US$1) million, up 40% in three years, and the shop floor kept saying, 'Not one line of this covers our equipment.' The talent development team spent four months weighing whether to bring training in-house.
- The head of talent development rebuilt the cost sheet. Next to the KRW 170 million (≈ US$127,200) outsourcing fee, they added the 640 hours a year that eight internal instructors would spend on preparation, converted to labor cost. The in-house option came out not at KRW 90 million (≈ US$67,300) but at KRW 140 million (≈ US$104,700).
- When the 26 training topics were laid out in a table, they split cleanly. The 11 on equipment operation, quality standards, and safety procedures covered content that outsiders could not create. For the 15 on leadership and report writing, everyone agreed they are the same wherever you learn them.
- The production engineering deputy managers named as internal instructor candidates asked, 'If we stop the line to go teach, who is accountable for that time?' HR responded by first writing a policy that counts one hour of teaching as one hour of recorded work.
- The plant manager was worried about inconsistent quality, so three candidates each gave a 30-minute trial lecture. Final approval came on the condition that the master course materials be kept on the team server, not on an individual’s laptop.
- The conclusion was a split. The 11 equipment and safety courses moved to internal instructors, while the 15 general courses stayed outsourced, with competing quotes from two vendors used to renegotiate unit prices.
- A year later, outsourcing fees had dropped to KRW 82 million (≈ US$61,300). Even after adding KRW 51 million (≈ US$38,200) in internal instructor operating costs, the total came in 21% lower. On-the-job application scores for the equipment courses rose from 54% to 79%.