The assumptions section is the tell

The assumptions section is the tell

The statementofwork.org thesis: no one has become a perfectly productive company by adopting a better template - but it's a start. Generative AI has made SoW contracts even more customised, and that broad creativity is not helping anyone. Adopting a standard, machine-readible, architecture of a SoW gets everyone on the same page.

Almost every bespoke SoW carries an assumptions section. The standard does not, and the absence is diagnostic.

An assumption is a sentence that says: if this turns out to be untrue, none of the above applies. It is drafted by the party who will rely on it and agreed by the party it will harm. "We assume timely access to stakeholders." "We assume the existing system is documented." Each reads as housekeeping. Each is a door left open at the precise point the document should be load-bearing.

The section exists because of how the document was made. When you write a scope fresh, you cannot foresee every dependency, so you hedge, and the assumptions list is where the hedges collect. It is the visible residue of a document that was invented rather than assembled, the place a novel SoW admits what it could not pin down.

Take the section apart and nothing in it is actually an assumption. Timely stakeholder access is a customer dependency: it has an owner, a date, and a stated consequence if it slips. System documentation is scope: either the supplier produces it or the buyer provides it. Move each line to where it can be enforced and the section empties. Everything in it was a commitment or a requirement in softer language.

A structure built from approved components does not accumulate hedges, because the components already carry their own dependencies. There is nowhere for an assumption to form.

The instinct behind most weak SoWs is to write each one fresh. Every freshly authored scope is a novel document that no one has reviewed, priced, or tested against delivery, and that novelty carries a commercial cost the buyer rarely sees until it arrives. The standard inverts the instinct. It begins from a baseline the enterprise has already approved, eleven sections and fifty-one elements, and permits variation only where the work genuinely differs. The schema turns that structure into data, the substrate AI needs to extract obligations, score risk, and govern a portfolio, rather than generate one more unreviewed draft. An assembled document has no assumptions section, because every dependency already has a field, an owner, and a date.

The launch

statementofwork.org launches on 10 August at the World Commerce and Contracting APAC Summit in Sydney. The full standard and engagement types, and the first learning and certification modules go live the same day. You signed up before any of it was public, which means you see it first.

Subscribe to Boring Statements of Work

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.
jamie@example.com
Subscribe