Statement of Work (SOW) Negotiation Playbook (2026)

What a statement of work is and how it sits with an MSA A statement of work (SOW) describes the specific work to be done under a services relationship: what will be delivered, by when, for how much, and how you will know it is done. It usually sits beneath a master services agreement (MSA), which holds the legal terms (liability, IP, confidentiality, warranties) that apply across all engagements. The split is deliberate: negotiate the MSA once, then each SOW should cover the work, not re open the legal terms. This playbook covers the SOW terms worth getting right, with positions and fallbacks. It is general guidance, not legal advice. For a material engagement, have a qualified professional review both the MSA and the SOW. ## Scope and deliverables: the heart of the SOW Most SOW disputes are scope disputes. Precision here prevents them. - Deliverables: list each one specifically, with a clear definition of done. Vague deliverables (improve performance) invite disagreement; specific ones (reduce page load to under two seconds on the listed pages) can be tested.

• Out of scope: state what is not included. Naming the exclusions is as valuable as naming the inclusions.
• Assumptions: record the assumptions the estimate relies on, so a broken assumption routes to change control rather than a free extra. ## Acceptance criteria and process - Criteria: objective, testable criteria for each deliverable. If it cannot be tested, it cannot be accepted or rejected cleanly.
• Process: a review window, the right to reject with reasons, a cure period to fix, and what happens if it still fails. Customer position: deemed acceptance only after you have actually had a fair chance to test. Supplier position: deemed acceptance after a set number of days to avoid open ended limbo. Fallback: a defined test window with deemed acceptance after it, and a clear re test loop. ## Pricing model and payment - Fixed price: best when scope is well defined; the supplier carries overrun risk, so expect a firmer change control stance.
• Time and materials: best when scope is uncertain; the customer carries overrun risk, so negotiate a not to exceed cap and regular burn reporting.
• Milestone based: tie payment to accepted deliverables, not to the calendar, so payment follows value.
• Expenses: agree what is reimbursable, any cap, and the approval process. ## Change control Scope will move. A working change control process is what keeps that from becoming a dispute. Agree a simple route: a written change request, an impact assessment on price and timeline, and sign off by both sides before work on the change starts. No change proceeds on a verbal nod. This single mechanism prevents most SOW conflict. ## Timeline, dependencies and responsibilities - Timeline: key dates and milestones, with float where the work is uncertain.
• Customer dependencies: what the supplier needs from you (access, data, decisions, people) and by when. Make relief from delay conditional on the supplier flagging a missed dependency promptly, so neither side hides behind it.
• Roles: named leads on each side, a governance cadence, and an escalation path. ## IP, warranties and the MSA link The SOW should say who owns the deliverables and any pre existing materials, but the detailed IP, warranty, liability and confidentiality terms should live in the MSA and simply be referenced. Watch for a SOW that quietly changes the MSA terms, for example a new liability cap or IP position buried in the work document. If the SOW and MSA conflict, the agreement should say which wins; usually the MSA governs legal terms and the SOW governs the work. ## Preparation and common traps Pin the scope before you argue price, because price without defined scope is meaningless. Decide your must haves (testable acceptance, a real change control process, and a dependency regime that is fair both ways). Common traps: vague deliverables, no out of scope list, deemed acceptance with no fair test window, a time and materials deal with no cap, and legal terms smuggled into the SOW that contradict the MSA. ## One workspace for draft, review, track and govern A statement of work is where good legal terms in an MSA either hold or quietly get rewritten, and where almost every services dispute actually starts. The deeper problem is that this work is usually split across a drafting tool, a review tool, a spreadsheet of dates, and email threads, so nothing stays connected and knowledge lives in people's heads. A single AI workspace tuned to your company takes a different approach. You draft, understand, edit, track and govern every contract in one place, where each document is modelled as a structured clause tree rather than flat text, so the system can show per clause risk, flag where a draft departs from your standard positions, let you edit a clause in plain language, and keep a live timeline of what changed and when. Because it learns from your own documents and preferred terms, the output sounds like your company rather than a generic template. Adira is built around that idea. A single-pane view of upcoming obligations and overall legal health is on its roadmap. The point is not another tool to check, it is to make the judgement on this page a repeatable company standard instead of something each person re-derives on every deal.