How to Write a Scope of Work (+ Examples & Templates)

Sorry, there were no results found for “”
Sorry, there were no results found for “”
Sorry, there were no results found for “”

“Can you just refresh the website?”
To one person, that means five pages and cleaner layouts. To another, it means a blog migration, new landing pages, updated copy, and a logo they have wanted to redo for months.
A scope of work fixes that. It defines what the project will deliver, how the work will be done, when each part is due, what counts as approved, and what falls outside the agreement.
That clarity matters more than it looks. The CHAOS Report by the Standish Group found that only about 31% of projects finish on time and within scope.
Here’s the uncomfortable part: most of those misses aren’t a discipline problem. They’re a file problem. The scope of work gets signed, emailed, and filed, and from that moment, it starts drifting from the actual work. By week three, the document says five pages and the project says fifteen, and nobody can point to where the gap opened. The fix isn’t a longer SOW. It’s a scope that stays connected to the work it describes.
This guide shows you how to write a scope of work that is clear enough to price, plan, assign, approve, and manage after kickoff.
A strong scope of work answers four questions: what you’re delivering, how, by when, and what counts as done. The strongest scope isn’t the most detailed, but the one that remains truthful and transparent after 30 days.
Details may fade once work starts, so having an up-to-date document throughout the project is essential. This ensures “out of scope” is clearly defined and referable.
Cover all seven components, and instead of vague verbs like “manage,” specify numbers such as “4 posts/month.” Spell out what you’re not doing, name the client-side dependencies that cause most delays, and always get written approval for every change request before starting the work. Then keep the signed doc connected to the live tasks, so when scope creeps, the change is visible and priceable instead of silent and free.

A scope of work (SOW) is a document that defines exactly what a project will deliver, how, by when, and what constitutes finished work. It clearly defines the scope of the work so that all parties—clients, vendors, and internal teams—agree on the deliverables before beginning.
Its job is alignment. The SOW names the objectives, lists the deliverables, breaks them into tasks, ties them to a timeline, and states how each deliverable will be judged complete. It becomes the reference everyone points back to when a question or a change comes up. That’s also why it doubles as the backbone of project scope management.
A scope of work answers what will be done; a statement of work answers how the two parties will work together to do it. The two share the acronym SOW and often get used interchangeably, but they operate at different levels.
A third similar term is project scope. This defines the boundaries of the project itself, what’s included and what’s not, rather than the work or the working relationship.
| Document | What it answers | Typically contains | When you use it |
|---|---|---|---|
| Scope of work | What work will be done | Objectives, deliverables, tasks, timeline, acceptance criteria, exclusions | Defining a specific engagement or project phase |
| Statement of work | How the parties will work together | The scope of work plus legal terms, payment schedule, warranties, governance | The full contractual agreement, often with the SOW nested inside it |
| Project scope | The full boundary of the project | All deliverables and the work required to produce them, plus what’s out of bounds | Internal planning and scope management across the whole project |
In practice, the statement of work is the umbrella contract, and the scope of work is the section inside it that spells out the deliverables. Project scope is the planning concept that both documents describe. If a stakeholder asks for “the SOW,” it’s worth a ten-second check on which one they mean, because the legal and payment terms live in the statement, not the scope.
A complete SOW should make the work easy to understand, price, plan, and approve. Miss one core component, and you leave room for confusion later.
Here’s what to include:
Scopes of work usually fall into four structures. The right one depends on how predictable the project is, how clearly you can define the deliverables, and what kind of risk each side is willing to take on.
Quick pick:
Most scope disputes start when the structure does not match the work. A deliverable-based SOW can work well for a fixed website build, but it can become risky for a project that still needs research, strategy, or stakeholder alignment. If you cannot name the deliverables yet, do not pretend you can in the SOW. Use a structure that reflects the uncertainty.
To write an SOW, define the objective, list deliverables and non-deliverables, break them into tasks, set a timeline with dependencies, write acceptance criteria, state assumptions and a change-control process, then add cost and sign off. These seven steps move a project from kickoff to closeout. In practice:
Start with why the work exists and what a good outcome should look like. This is the anchor for the entire SOW. If the objective is vague, the deliverables, timeline, and approval process will be vague too.
Write the objective in one or two sentences that the client would recognize. Avoid internal jargon and broad phrases like “improve brand presence” or “support marketing efforts.” Instead, tie the work to a clear business outcome.
For instance:
Then add success criteria. This does not always need to be a hard KPI, especially if the project is creative or discovery-heavy. But it should explain how both sides will know the work did its job.
Next, name every output the client will receive. Be specific, countable, and forthright.
Avoid broad verbs like “manage,” “support,” “handle,” or “optimize” unless you define what they include. Those words sound useful in a proposal, but they create room for scope creep inside an SOW.
For example:
The stronger version tells the client what they are getting, how much they are getting, and where the boundary is. Then list the non-deliverables in the same section. This is where you spell out what is not included in the work, even if it feels obvious to you.
For a blog project, non-deliverables might include:
Denver International Airport’s automated baggage system is a useful warning for any SOW.
The headline deliverable sounded simple: build an automated baggage system for the new airport. But the actual work involved multiple concourses, airlines, baggage types, routing rules, and operational exceptions.
According to a GAO review of the project, Denver awarded BAE Automated Systems a contract worth about $195.6 million in 1992. By 1995, the cost had risen to over $290 million. The airport’s opening was also delayed from October 1993 to February 1995 due to system problems and major modifications.
The scope kept changing through added conveyors, odd-sized baggage handling, maintenance equipment, routing updates, and airline-requested changes. One later agreement reduced baggage-handling capacity from 65 bags per minute per line to 30. Denver also had to build a conventional backup baggage system, which GAO estimated at $63 million.
The Lesson: Don’t define only the big deliverable; define the boundaries around it, too. “Automated baggage system” should have named which airlines, baggage types, exceptions, fallback processes, and testing standards were in or out.
But boundaries on paper weren’t the whole fix. Denver had a scope. What it didn’t have was a scope connected to live execution, so every conveyor, routing change, and airline request landed as a side agreement that the original document never absorbed. The file and the build became two different projects.
After you list each deliverable, run a simple test: could someone start work from this line tomorrow?
If the answer is no, it is still too broad.
Take one deliverable at a time and break it into task-level actions using a verb + object format. For example, “create homepage wireframe,” “review legal copy,” “approve final design,” or “upload blog post to CMS.” Avoid vague task names like “website work,” “content support,” or “design updates.” They do not show who does what.
For a blog deliverable (4 SEO blog posts per month), the task breakdown might look like this:
| Task | Owner | Dependency |
|---|---|---|
| Confirm topics and keywords | Content strategist/SEO strategist | Client approves content plan |
| Create briefs | Content strategist | Keywords confirmed |
| Write first drafts | Writer | Briefs approved |
| Review for product accuracy | Client SME | Drafts submitted |
| Edit for structure and clarity | Editor | SME comments added |
| Add internal links and metadata | SEO specialist | Final edits complete |
| Upload to CMS | Content manager | Client approves final draft |
This task layer exposes the real workload behind the deliverable. It also shows where delays can happen. If the client SME does not review the draft on time, the editor, SEO specialist, and content manager all get pushed back. That dependency should be visible before the project starts.
For larger projects, use a work breakdown structure. This means grouping the work by phase, then breaking each phase into assigned tasks. For example, a website redesign might include discovery, sitemap, wireframes, copy, design, development, QA, analytics setup, and launch. Each phase should then become task-level work with an owner, due date, and approval point.
You can build this task map in:
Now turn the task list into a timeline. Add the start date, end date, major milestones, and any dependencies that could affect delivery.
Do not only write “Project due in six weeks.” Break the timeline into phases, such as discovery, first draft, review, revisions, final approval, and launch. Then mark which milestones trigger payment, approval, or the next phase of work.
For example:
The most important part is naming dependencies, especially the ones on the client’s side. Many project delays do not come from the actual work. They come from late feedback, missing login access, delayed legal reviews, unavailable stakeholders, or assets that arrive after the team has already started.
Spell these out before work begins.
For example: “Project timelines depend on timely client feedback, access to required tools, and delivery of approved brand assets. Delays in feedback, approvals, access, or materials may shift the project timeline by the same number of business days.”
For each deliverable, define what counts as approved. This is the clause that stops a project from drifting into endless “almost there” revisions.
Keep the criteria tied to observable checks. For a website migration, “launch-ready” could mean redirects are tested, analytics are firing, contact forms work, and the approved pages load correctly on desktop and mobile. For a sales deck, it could mean something completely different.
Avoid approval language that relies solely on taste, such as “the client is happy with the work.” A better line would be:
“Deliverable is accepted once it meets the agreed criteria, includes the approved revision round, and receives written sign-off from the client.”
This gives both sides a clean closing point. New requests after that can still occur, but they are handled as change requests rather than extending the original scope.
This is where you protect the project from quiet expansion.
That last part matters most. If the vendor completes extra work first and discusses payment later, the change request becomes much harder to enforce.
The FBI’s Virtual Case File project is a strong example of why change control and milestones need to be written into the scope early.
The project was meant to modernize the FBI’s case management system. But according to a DOJ Inspector General review, the FBI and its contractors lacked a firm understanding of the design requirements as the work progressed. One FBI project manager said the Trilogy program’s scope grew by about 80% after the project began.
The contract structure also made the problem harder to control. The review found that the statements of work did not include specific completion milestones and critical decision review points, and did not set penalties if milestones were missed.
The project was eventually abandoned after about $170 million had been spent.
The Lesson: Write the change process before the work starts. The SOW should specify how changes get requested, priced, approved, and folded into the timeline, with real milestones and review points wired in. Otherwise, scope changes become informal decisions, and informal decisions become expensive.
It’s the same failure as a filed PDF, just at a $170M scale: the statements of work were static while the requirements moved. With nothing connecting the scope to the live work, there was nothing to catch the 80% growth as it happened.
End the SOW with the commercial terms: total price, billing structure, payment schedule, due dates, late-payment terms, and the person authorized to approve the work.
Tie payments to clear milestones where possible. For example:
For ongoing work, state the monthly retainer, billing date, scope included, and what happens when the client exceeds the agreed-upon hours or deliverables.
Then get a written sign-off. An SOW without approval is just a working draft. Signatures confirm that both sides agree on the deliverables, timeline, exclusions, payment terms, and change process before the work begins.
The Scope of Work Template by ClickUp gives you a ready-made Doc for capturing project details, deliverables, responsibilities, timelines, change management, budget, and approvals in one place. It is useful when you want the SOW to become part of the project workflow.
Why use this template:
An SOW uses the same core sections across industries: objectives, deliverables, timeline, assumptions, exclusions, acceptance criteria, payment terms, and sign-off.
What changes is the risk.
A website project usually fails around page count, revisions, CMS ownership, and post-launch support. While a construction scope fails around permits, site conditions, inspections, and material changes.
Use the examples below as guides to shape your own scope, not as contracts to copy word-for-word.
Here’s a website redesign SOW in miniature, with every core section filled in:
Objective: Replace the marketing site to help B2B buyers understand the product faster and lift demo signups.
Deliverables: 8 designed and developed pages based on the approved sitemap, 2 revision rounds per page, CMS handoff and a 1-hour training session.
Timeline: 6 weeks across 3 milestones, design sign-off (week 2), build complete (week 4), launch (week 6).
Acceptance criteria: Each page is approved against the signed design comp, loads correctly on desktop and mobile, and receives written client sign-off.
Exclusions: Copywriting, stock photo licensing, new pages beyond the sitemap, SEO migration, post-launch maintenance.
Payment: 40% at signing, 30% at build complete, 30% at launch.
Change control: Any request outside this scope is logged, priced for cost and timeline impact, and approved in writing before work starts.
The risk in website work is hidden labor. A client may assume copy, SEO migration, new pages, stock photos, custom graphics, CMS upload, and post-launch support are all included unless the SOW says otherwise.
Include:
Exclude:
Sample SOW language:
“Vendor will design and develop eight website pages based on the approved sitemap, with up to two revision rounds per page. Copywriting, stock image licensing, additional page requests, SEO migration, and post-launch maintenance are excluded unless approved through a written change request.”
The risk in construction is assuming the “build” includes everything around the build. Permits, inspections, site access, weather delays, material cost increases, and owner-driven design changes need explicit treatment.
Include:
Exclude:
Sample SOW language:
“Contractor will complete site preparation, framing, electrical rough-in, and finish work according to the approved drawings and specifications. Permit fees, owner-requested changes, unforeseen site conditions, and material upgrades are excluded and will be handled through a change order.”
The risk in software is vague feature language. “Build dashboard” can mean ten different things to ten stakeholders. Define the feature, test condition, environment, integration responsibility, and post-launch support window.
Include:
Exclude:
Sample SOW language:
“Vendor will develop the user stories listed in Appendix A and deploy them to the staging environment for client review. Features outside Appendix A, third-party subscription fees, historical data cleanup, and post-warranty support are excluded unless approved through a change request.”
The risk in marketing lies in conflating deliverables with outcomes. You can scope campaign strategy, assets, reporting, and launch support. You should be careful about promising leads, revenue, CAC, or ROAS unless the vendor controls the media budget, landing page, sales follow-up, and tracking setup.
Include:
Exclude:
Sample SOW language:
“Vendor will provide campaign strategy, three ad creative concepts, landing page copy, two marketing emails, and a performance reporting dashboard. Paid media spend, influencer fees, additional creative variants, and sales follow-up are excluded. Performance targets are directional unless separately tied to approved budget, tracking, and campaign controls.”
The risk in retainers is unclear capacity. “Ongoing support” can become unlimited calls, strategy, execution, reporting, and ad hoc requests unless the SOW defines hours, response times, meeting cadence, and rollover rules.
Include:
Exclude:
Sample SOW language:
“Vendor will provide up to 20 hours of consulting support per month, including one weekly advisory call, async review of agreed materials, and a monthly summary of recommendations. Unused hours do not roll over. New projects, same-day requests, and work beyond 20 hours require written approval.”
The useful move is the same across every industry: count the deliverables, name the common assumptions, and write exclusions around the requests people are most likely to make later.
By now, the pattern is clear: a signed PDF can’t track a moving project. So the real question isn’t how to write a better SOW, it’s what the SOW becomes after the signature.
Treat it as two things, not one. The signed agreement stays frozen as the baseline, the record of what both sides agreed to at the start. The work around it stays live: tasks, owners, timelines, approvals, budgets, risks, and change requests. When those two are connected, a new request can’t quietly become free work; it shows up against the baseline as a change.
Here’s the difference:
| Static SOW | Working SOW baseline |
|---|---|
| Lives as a PDF or document attachment | Connected to tasks, timelines, approvals, and budget |
| Reopened only when something goes wrong | Referenced during delivery and review cycles |
| Changes happen through Slack, email, or side calls | Changes are logged, priced, approved, and tied to timeline impact |
| The task list drifts away from the original agreement | Execution stays traceable to the approved scope |
The signed SOW should not be edited casually every time something changes. This is the work-sprawl problem in miniature, where the update lands in a DM, the SOW sits in a separate document, and the task list lives somewhere else.
This is where tools like ClickUp help. The SOW can live beside the tasks, comments, approvals, deadlines, and budget.
This short ClickUp clip shows what that looks like and how teams pull it back together.
Best Practice: Always keep the original version stable and manage changes through a clear change log or approved change request. For example, if a client asks for two extra landing pages during a campaign, the request should not simply appear on the project board. It should be logged as a scope change, reviewed for cost and timeline impact, approved in writing, and then added to the work plan.
An SOW is easier to manage when the document, tasks, approvals, and timeline stay connected:

Start with the SOW in ClickUp Docs. You can use the Doc to capture the objective, deliverables, exclusions, timeline, acceptance criteria, and payment terms. Then, as the project progresses, the Doc can remain connected to the tasks and milestones it governs.
For example, the “Deliverables” section of a website SOW can link to tasks for homepage copy, wireframes, design review, development, QA, and final approval. That makes the SOW easier to reference during delivery.
Once the SOW is written, turn each deliverable into a task or group of subtasks with ClickUp Tasks.
For example, “three landing pages” can become separate tasks for each page. Each page can then have subtasks for copy, design, development, review, QA, and approval.
This helps the team see the real work behind each deliverable. It also reduces the risk of silent scope expansion. If a fourth landing page appears on the task list, it is easier to spot because it does not match the original SOW.
ClickUp Dashboards help when you are managing multiple SOWs at once.
An agency, contractor, or operations team can use dashboards to track overdue approvals, work by client, upcoming milestones, open change requests, workload, and budget signals. This makes scope risk easier to catch before it becomes a margin problem.
For example, if three client projects are waiting on feedback, a dashboard can show that delay across accounts instead of burying it inside separate task lists.
ClickUp’s Dashboards changed how we run day-to-day agency operations. We monitor capacity across three dev teams and surface blockers before they become delays, so we spend less time chasing status in Slack threads and more time actually moving client work forward.
ClickUp Brain can draft an SOW from a project brief, summarize lengthy client calls, turn notes into slides, or answer questions using team data from Tasks, Docs, Chat, and connected workspaces.
For SOW, that means you can ask questions like:
Brain also expands this context-led direction further, with AI built around the company’s projects, docs, people, conversations, and knowledge.
Super Agents are more useful once the process is already defined.
For example, a team could set up an agent to prepare a weekly scope review, summarize overdue approvals, flag tasks missing a scope status, or draft a client update before a milestone meeting.
A good use case would be:
“Every Friday, summarize all open change requests, tasks added this week, delayed approvals, and milestones at risk for this client project.” That gives the project manager a faster review loop. It, however, should not replace sign-off. Any extra work still needs written approval before it becomes part of the scope.
ClickUp is useful when scope drift has real cost: agencies, service teams, contractors, consultants, operations teams, and internal teams managing several projects at once.
For a solo freelancer writing a one-page SOW for a simple project, it may be more structure than needed. A clean document, a shared task list, and a signed approval may be enough.
The payoff appears when the SOW is no longer just a document. It becomes the baseline your team can connect to tasks, timelines, approvals, change requests, and client communication.
| Mistake | Why it causes problems | Fix |
|---|---|---|
| Vague deliverables | Words like “manage,” “support,” or “optimize” can be interpreted in several ways when no quantity or boundary is attached | Name the output, add a number, and define the handoff |
| No exclusions list | Anything left unsaid can be assumed to be included, especially in client-service work | Add an out-of-scope section for related work that the client might reasonably expect |
| Counting revisions without defining a revision | “Two rounds” can still become messy if one round includes ten unrelated changes from five stakeholders | Define what one revision round includes, who can request it, and when it ends |
| No change-control process | Every new request becomes an informal negotiation once work has already started | Require written approval for any scope change, including cost and timeline impact |
| Ignoring client-side dependencies | Late feedback, missing files, or delayed approvals can push the timeline while the vendor absorbs the blame | State what the client must provide, by when, and how delays affect delivery dates |
A good SOW pays off three times over. Writing it forces the fuzzy parts to get specific. Running the project, it settles disputes before they escalate. And when scope creeps, it makes the change visible instead of letting it slip through for free.
But all of that value rides on one condition: the document can’t die in an inbox.
The teams that get the most from it treat the scope as a living system: defined clearly up front, connected to the actual tasks, and revised as the work evolves. Cover the seven components, quantify every deliverable, spell out the exclusions, and keep the document tied to the work it describes.
The fastest way to put this into practice is to start from a template and connect it to your project. Get started with ClickUp for free, customize the SOW Template in Docs, then link each deliverable to the tasks that deliver it.
A scope of work is a document that specifies the deliverables and tasks for a specific engagement or phase. Project scope is the broader planning concept that defines the project’s full boundary: everything included and everything excluded. In short, the scope of work describes a slice of the work, while project scope describes the entire boundary that the work fits inside.
An MSA sets the overarching legal and commercial terms governing an ongoing client-vendor relationship; a scope of work defines the deliverables for one specific project under it. The typical hierarchy is MSA at the top, individual statements of work beneath it, and the scope of work nested inside each statement. The MSA is signed once; a new SOW is issued for each engagement.
The four common structures are deliverable-based (payment tied to specific outputs), time-and-materials (billing for hours and costs as the work happens), level-of-effort (a set capacity over a period, common for retainers), and performance-based (payment tied to measurable outcomes). Choose the structure that matches how predictable the deliverables are. Most scope disputes start when the structure doesn’t fit the work.
The party doing the work usually drafts the scope of work: the agency, vendor, or internal project lead. Both sides then review and sign off before work begins. Drafting it yourself is an advantage because it lets you set the deliverables, boundaries, and exclusions on your terms.
A scope of work becomes legally binding once it’s signed as part of a contract or statement of work. On its own, it primarily defines the work, but inside a signed agreement, it becomes the reference for what was promised. For anything contractual, have the binding terms reviewed by a qualified professional.
Yes. A scope of work operates between internal teams—marketing commissioning a data dashboard, or ops requesting an internal tool—exactly as it does between a client and a vendor. The components are identical: objectives, deliverables, timeline, acceptance criteria, and exclusions. The only difference is that sign-off comes from an internal stakeholder instead of an external client, and payment terms may be replaced by budget or resource allocation.
A scope of work is sometimes called a statement of work, a project scope statement, or simply “the scope,” though these aren’t perfect synonyms. A statement of work is the broader contract; a project scope statement is the internal planning version. If someone uses the terms loosely, confirm whether they mean the deliverables (scope of work) or the full agreement (statement of work) before you act on it.
A scope of work states what will be delivered and the terms around it; a work breakdown structure breaks those deliverables into a hierarchy of sub-deliverables and tasks. The SOW is the agreement; the WBS is the execution map underneath it. You build the WBS from the SOW’s deliverables, which is how the document stays connected to the actual tasks instead of drifting.

Praburam Srinivasan
Max 24min read

Jeremy Galante
Max 21min read

Manasi Nair
Max 18min read

© 2026 ClickUp