When to Build a Team and When to Buy Capability

Build-vs-buy conversations almost always mean software. There’s a parallel, less-discussed version of the same question for capability itself: hire a team, or buy access to one through a contractor, agency, or fractional arrangement. Getting this decision wrong is at least as expensive as the software version, and it gets considerably less deliberate attention.

I’ve written about the general build-vs-buy framework for software. The team-versus-capability version needs its own distinct test, because the trade-offs run differently — a full-time hire isn’t a system that can be swapped out cleanly, and an external capability relationship carries genuinely different risks than a vendor contract does.

The test that actually discriminates

Is the need durable or temporary? A capability the business will need continuously, for years, justifies the investment of building an in-house team — the fixed cost amortises over a long enough horizon to be worth it. A capability needed intensely for a defined period — a licensing push, a migration project, a specific regulatory deadline — rarely justifies permanent headcount, because the team would sit underutilised the moment the specific need passes, and disbanding a team is organisationally and humanly harder than not building one you didn’t need permanently.

Does the capability need deep institutional context, or portable expertise? Some capabilities genuinely require deep, accumulated knowledge of the specific business — its history, its idiosyncrasies, its relationships — that an external party can never fully replicate no matter how skilled. Others require genuine expertise that’s largely portable across organisations, where an external specialist’s broader pattern-matching across many similar situations is actually more valuable than deep institutional context would be.

Can the organisation actually manage what it’s buying? Buying external capability still requires internal capability to direct it well — someone who can specify the work, evaluate the output, and hold the external party accountable. An organisation that buys expertise it has no internal capacity to manage is often paying for capability it can’t actually direct effectively, which is a frequently underestimated cost of the “buy” option specifically.

What does reversibility actually cost in each direction? Undoing a hiring decision is a genuine, difficult, human process — redundancy, morale impact, institutional disruption. Undoing an external relationship is comparatively straightforward — a contract ends, a new arrangement begins. This asymmetry matters most when the underlying need is itself uncertain, because the option that’s cheaper to reverse carries less risk if the initial judgment turns out to be wrong.

Where the fractional model specifically fits

A fractional arrangement — the model I work under myself — sits deliberately between full-time hiring and pure contracting, and it earns its place specifically where the need is genuinely durable (ruling out a short-term contractor) but doesn’t require full-time capacity, or where the organisation needs senior judgment applied continuously without committing to the reversibility cost of a full-time senior hire before the need is proven out. This connects directly to the total-cost comparison I’ve written about separately — the right model depends on matching the actual shape of the need, not defaulting to whichever model is most familiar.

Where the test actually changed the decision

A growth-stage logistics firm defaulted toward hiring a full-time head of engineering, the familiar move for a company at its stage. Running the actual test surfaced a different picture: the need was durable, but the firm’s actual gap was senior architectural judgment applied a few days a week, not full-time engineering management — a distinction the firm hadn’t separated clearly until asked directly. A fractional arrangement, reversible within a quarter if the fit was wrong, filled the actual need at a fraction of the commitment a full-time hire would have required, with the option to convert to a full-time search later once the shape of the role had proven out through real engagement rather than a job description written in the abstract.

Working through whether a specific capability need is best served by hiring, contracting, or a fractional arrangement — matched to the actual shape and duration of the need — is exactly the kind of scoping conversation a fractional CTO engagement starts with.

The test is worth applying honestly even when the answer points away from the model currently being considered — a fractional arrangement isn’t the right answer to every capability gap, any more than a full-time hire is.

Getting this decision wrong in either direction is expensive — an unnecessary permanent hire for a temporary need, or a string of short-term contractors for a need that genuinely warranted institutional continuity.

Where the answer points toward fractional rather than full-time, the true cost comparison is worth running explicitly.

The honest version of this conversation is uncomfortable for anyone with an interest in one particular answer — including, worth naming directly, for a fractional consultant whose own model isn’t always the right fit.