Insourcing vs Outsourcing: A Project Governance Decision

The question usually arrives already framed as arithmetic. Someone has a supplier day rate, someone else has an internal cost recovery figure, and the two are set beside each other as though the answer falls out of the comparison. It rarely does. By the time a project is deciding whether to build a workstream internally or buy it in, most of what matters has stopped being about the difference between two numbers.

A more useful framing is this. Sourcing decides who does the work, who decides how it is done, who answers for it when it goes wrong, and what the organisation can still do once the work is finished. Only the first of those four is a procurement question.

As a short answer: keep work internal when the organisation needs to hold the capability afterwards, when accountability for the outcome cannot realistically be delegated, or when the work is so entangled with local knowledge that specifying and supervising it would cost more than doing it. Buy it when you need capability you do not have and do not need to keep, when the work is bounded enough to specify and accept, or when a supplier genuinely carries a risk better than you can. The interesting projects are the ones where different parts of the same workstream answer differently.

Why sourcing sits inside governance

The PMBOK® Guide Eighth Edition treats insourcing and outsourcing considerations as part of the Governance Performance Domain at Section 2.1, alongside project authority, oversight, integrated planning and change. That is a sensible place for it, because a sourcing decision moves three things and leaves one exactly where it was.

Capability moves. Whoever does the work becomes more fluent in it, and whoever does not, does not. Some risk moves, though usually less than the contract implies: a fixed price transfers the cost of overrun, but it does not transfer the consequence of late delivery to a regulator, a customer or an operating site. Decision rights move partially, because the supplier decides method inside whatever boundary the specification draws. Where the specification is thin, more decision-making has been handed over than anyone intended.

Accountability does not move at all. The sponsor still answers to the organisation for the outcome, and the project manager still answers for the schedule. Giving work to a supplier changes who performs it, not who is answerable when it is wrong, and governance exists partly to keep that line visible when a contract makes it look blurred.

There is also an arithmetic point that day-rate comparisons routinely miss. Outsourced work creates internal work: specification, tendering, mobilisation, access and escorting, acceptance, invoice checking, commercial management, and the ordinary supervision of people who do not yet know your organisation. Internal effort changes shape rather than disappearing, and on a small workstream it sometimes does not fall at all.

Capacity and capability are different problems

Most sourcing decisions are triggered by one of two pressures, and the two are frequently confused.

A capacity shortfall means you know how to do the work and do not have enough people to do it inside the window. It ends when the peak ends. A capability gap means you do not know how to do the work, or not to the standard required, and it does not close when the project does unless somebody deliberately closes it.

The confusion produces two recognisable failure modes. The first is outsourcing a capability the organisation needed to retain, because the immediate problem presented as a capacity peak. The work is delivered, the project closes, and some months later nobody internally can maintain, modify or defend what was built. The second is insourcing a genuine capability gap because the internal figure looked lower. That buys a slower, riskier delivery and a team learning in public on a live project, which is occasionally the right call and is almost never priced as one.

Cost belongs in the decision. It is simply not the first question, and cheapest is not automatically correct. Strategic importance, intellectual property, the health of the supplier market and the long-term operating implications all sit alongside it.

Development approach matters here too, because what you can safely buy depends on what you can specify. Predictive work with stable, well-understood scope contracts cleanly, since acceptance criteria can be written before the work starts. Adaptive work does not, and organisations that try to buy an evolving product against a fixed scope tend to end up managing a change log rather than a product. Where delivery is iterative, augmenting an internal team is often more honest than contracting for an outcome nobody can yet describe. Hybrid projects frequently split the difference: the specifiable elements are bought, the evolving ones stay in-house.

Splitting a qualification workstream

A life sciences site is installing a new packaging line. The qualification workstream, covering protocols, execution, documentation and deviation handling, is estimated at around nine months of effort at peak. The site's validation team is four people who already carry routine requalification duties for the existing lines. Recruiting takes months, and the additional roles would be surplus once the new line is running.

The obvious move is to buy the workstream from a validation consultancy that has done it many times before. The obvious counter-move is that the site's quality unit, not the consultancy, is the party that answers to the regulator.

The project manager separates the work that ends from the work that continues. Protocol authorship and execution for the initial qualification is a peak: well understood, specifiable, and not recurring at that volume. It is bought. Approval of those protocols, decisions on deviations, and ownership of the change process that will requalify the line after every future modification stay internal, because the site will be doing that work for the next decade and because the accountability was never transferable in the first place.

Two consequences follow, and both belong in the plan rather than in the contract. Contractor time inside the classified area needs internal escorting and access approval, which is internal effort the day-rate comparison did not contain. And the internal team's involvement in approving the early protocols is not a governance formality, it is how the site acquires enough familiarity with the new line to run it afterwards.

Change one fact and the answer changes. If the line were being installed to run for two years ahead of a planned site closure, buying the whole workstream would be defensible wherever regulation allowed it. Retained capability has no value when nothing is being retained.

What to test before the decision is signed off

A short set of questions does most of the work here. They are not a framework, simply the things worth establishing before a recommendation reaches a sponsor.

• What must we still be able to do the day after the supplier leaves?
• Can we specify this well enough to accept it, and who signs that acceptance?
• Is this a peak that ends, or a gap that stays open?
• What internal effort does buying create, rather than remove?
• Which risks genuinely transfer, and which only appear to?
• What does the answer become if the volume doubles, or if the scope turns out to be less stable than assumed?

If those cannot be answered, the decision is not ready, whatever the cost comparison says. This is the kind of judgement we work through in structured form during PMP® Exam Preparation, because recognising which question a situation is actually asking travels further than any single decision rule.

For a PMP candidate, the important distinction is between a sourcing question and a cost question. Preparation material can leave the impression that make-or-buy is settled by comparing internal and external cost, which is a simplification that works on a spreadsheet and struggles in a governance forum. The useful exam-preparation habit is to read a sourcing scenario for what it says about specification, acceptance, retained capability and who answers for the outcome, then let the cost comparison inform a decision it cannot make on its own.

On a real project this surfaces early, usually before anybody has used the word governance. It appears as a resourcing plan with a hole in it, or a business case with a supplier quotation already attached. The project manager who notices that the decision will outlive the project, and says so while it is still open, is doing considerably more useful work than the one who reconciles the two numbers accurately.

Andre Malowney

Interested in going further?

Sourcing decisions are seldom difficult because the arithmetic is hard. They are difficult because specification, accountability and retained capability have to be weighed together, usually with a delivery date already fixed. Structured PMP® preparation gives you a way to hold those considerations in view at once, and to recognise the same pattern when it appears in scope, risk and stakeholder decisions.

The Governance Performance Domain in the PMBOK® Guide Eighth Edition is where sourcing sits alongside authority, oversight and change, and it is the natural place to read further on how those connect.

Ad · Amazon affiliate link.

References

A090: Project Governance Explained: Who Decides What, and Why
A091: Project Governance vs Organisational Governance: What's the Difference?
A092: How Projects Create Value Through Governance
A093: Choosing a Project Governance Model
A094: Governance Metrics That Actually Tell You Something

PMP and PMBOK are registered marks of the Project Management Institute, Inc.