A statement of work buys an outcome instead of hours: defined deliverables, milestones and acceptance criteria, with the supplier carrying the risk of reaching them. Managing it means keeping that structure real. We do that across 11 countries, where a statement of work that is really staff augmentation in disguise stops being a procurement problem and becomes a legal one.
Of managed workforce spend now runs through statements of work, against 18% in 2017. The fastest-moving line in external spend, and the least governed.
Of it is misclassified staff augmentation, on one provider's estimate. Priced as consulting, delivered as hours.
The Belgian law a disguised statement of work runs into. English-language SOW guidance rarely mentions it.
Scope and deliverable definition
Including what is explicitly out of scope.
Milestones and acceptance criteria
Agreed before the work starts, not after.
Supplier selection for outcomes
Different criteria from selecting for hours.
Change order discipline
Where scope creep gets priced, not absorbed.
Spend visibility
Statements of work in the same view as staffing.
Structure testing
Whether it survives a labour inspection.
Milestone-based payment
Because time and materials undoes the point.
Risk that moves
The point of buying an outcome is that the supplier carries the risk of delivering it. Written loosely and paid by the hour, that risk quietly stays with you while you pay a premium for having supposedly transferred it.
Spend you can actually see
Statements of work signed inside business units land in procurement or consulting ledgers rather than in your workforce reporting. Bringing them into one view usually surprises the person who signed off the budget.
Scope that survives contact
Deliverables, milestones and acceptance criteria written before anyone starts, with a change process for what comes after. Most disputes we see are scope arguments held after the fact.
Legally sound
The structure tested against the rules that apply where the work happens, not only against your procurement policy. In Belgium and France those rules reach the client, not only the supplier, which is the part that gets missed.
Governed SOW | Unmanaged SOW | By the hour | Consulting | |
|---|---|---|---|---|
| What you buy | A defined outcome | Nominally an outcome | Capacity, honestly labelled | An outcome, plus our team |
| Delivery risk | Supplier, and it holds up | Stated as supplier, actually yours | Yours, openly | Sparagus, fully |
| Payment basis | On milestones, on acceptance | Monthly, on time spent | On hours worked | On the engagement |
| Visibility | In one workforce view | In a procurement ledger | Good, if programme-managed | One engagement, one owner |
| Fits when | Recurring project spend across suppliers | Never, and it is the most common | You direct the work day to day | You want the problem taken away |
Send us three statements of work and how those teams are actually directed.
Spend discovery
Finding the statements of work already signed, across ledgers they were never reported in.
Classification review
Which engagements are genuine outcome purchases and which are hours wearing a different contract.
Structure and templates
Scope, milestone and acceptance templates that hold, plus the change order route and the approval thresholds.
Govern and review
New statements of work screened at intake, milestones checked against acceptance rather than against invoices, and renewals treated as decisions. Reviewed quarterly, because a clean structure drifts back one small renewal at a time.
Of managed workforce spend now runs through statements of work, against 18% in 2017. The fastest-moving line in external spend, and the least governed.
Of it is misclassified staff augmentation, on one provider's estimate. Priced as consulting, delivered as hours.
The Belgian law a disguised statement of work runs into. English-language SOW guidance rarely mentions it.
Software and application
Development work bought as a project, delivered as a team sitting in yours.
Data and business intelligence
Genuinely project-shaped work, which makes it the best candidate for a real outcome contract.
Cloud, DevOps and SRE
Migrations start as projects and end as run work nobody re-contracted.
Testing and validation
Easy to define, easy to accept, and often the cleanest statement of work on the estate.
Engineering studies
Deliverables are documents, so acceptance criteria can be written properly.
Business process work
Bought by the function rather than by procurement, so it rarely reaches reporting.

Financial services
External technology work now carries its own regulatory question on top of the labour one.
Life sciences and pharma
The sector with the highest share of managed spend running through statements of work.
Energy and utilities
Site work bought as a project, then supervised by the client, which is exactly the trap.
A contract to buy a defined outcome rather than a quantity of hours. It names deliverables, milestones, acceptance criteria and a price, and the supplier carries the risk of getting there. Buying by the hour is the opposite: you direct the work, you carry the risk, and you pay for time whatever it produces. Both are legitimate. Confusing them is where the trouble starts.
When the substance does not match the label. The signs are consistent: no measurable deliverable, payment on time spent rather than on acceptance, your managers directing the people day to day, and renewals that roll on with no change of scope. One large provider estimates more than a fifth of statements of work fall into this category. It is usually not deliberate. It is a manager routing around a tenure rule or a headcount freeze.
Because in most markets a disguised statement of work is a cost and control problem. Here it is also a liability one. Belgium prohibits hiring out workers where the user exercises part of the employer's authority, and a genuine service contract is what keeps you outside that. France prohibits for-profit arrangements whose exclusive object is supplying labour, and what saves you is real expertise and a defined result. English-language guidance on SOW governance almost never mentions either.
On this service, we govern. We help you define the scope, choose suppliers who can be held to an outcome, write acceptance criteria that mean something, and keep the structure honest over time. If you want us to carry the delivery instead, with our own team and our own accountability for the result, that is a consulting engagement and a different conversation.
On milestones, against acceptance. The moment payment tracks time spent, the risk transfer stops being real and the arrangement starts looking like hours to anyone examining it later. Where the work genuinely cannot be milestoned, that is a signal it is capacity rather than an outcome, and buying it by the hour openly is both cheaper and safer.
By looking outside workforce reporting. Statements of work usually sit in procurement or consulting ledgers, below approval thresholds, or inside supplier relationships nobody reads as workforce spend. We reconstruct the population from accounts payable rather than from the programme, which is why the first number tends to surprise people.
Send us three statements of work and a description of how those teams are actually directed. You get a written view of which ones hold, which ones do not, and what it is costing you.