Proposal vs Contract vs SOW: What to Send When

You send a proposal, the client replies "looks great, let's do it", and you start work. Three weeks later there is a disagreement about revisions, and you both scroll back through email trying to figure out what was actually agreed. Was the proposal the agreement? Was the "looks great" an acceptance? Does the revision limit you mentioned on the call count if it never made it into writing?

This confusion is not a niche problem. It is the default state of solo freelancing, because the three documents involved (proposal, contract, statement of work) have overlapping content, similar-sounding names, and wildly different jobs. Corporate teams have legal departments to keep them straight. You have this guide.

While researching this article we fetched the 4 top-ranking pages for "proposal vs contract" (PandaDoc, Better Proposals, Bonsai, and UpCounsel, July 2026). Our count: 0 of 4 mentioned the state freelance written-contract laws that now apply to millions of US freelancers, and 0 of 4 explained where the statement of work fits. That is the gap this guide closes.

The 30-second answer: three documents, three jobs#

A proposal persuades, a contract protects, and a statement of work defines done: three documents, and a typical solo project needs exactly two of them in writing.

Each document does one job:

• The proposal is a sales document. Its scope section should be written carefully (more on why below), but its job is to win the work, not to govern it.
• The contract is a legal document; it exists so that the worst-case scenarios (client never pays, client claims to own your entire portfolio, project dies halfway) have pre-agreed answers.
• The SOW is a definition document; it exists so that "done" is a checklist instead of a feeling. We cover its 9 sections in the statement of work guide.

One quotable rule: the proposal answers "why hire me", the contract answers "what if things go wrong", and the SOW answers "how will we both know this is finished".

Is a proposal legally binding?#

A proposal becomes legally binding when it meets the 3 formation elements of a contract: an offer, an acceptance, and consideration (something of value flowing both ways). Under US contract law, courts look for mutual assent to definite terms plus consideration, not for a document titled "Contract". The label at the top of the page does not decide anything; the content and the acceptance do. See Cornell Law School's contract overview for the formation elements.

In practice this cuts both ways for you:

It can protect you. If your proposal states the deliverables, the price, and the payment terms, and the client accepts it in writing ("yes, let's proceed" in an email can qualify), you likely have an enforceable agreement even though nothing was ever called a contract. This is why your proposal's scope and price should always be exact: it may end up being the operative document.

It can also trap you. If your proposal says "unlimited revisions until you are happy" as a sales flourish, and the client accepts the proposal, that flourish is now arguably a term. Anything you would not want enforced against you should not appear in the proposal.

And it can fail you when you need it. A proposal that leaves key terms open ("final scope to be discussed") or that the client never clearly accepted is not a contract, no matter how much work you have already done on the strength of a phone call. A signature added "for discussion purposes" does not create acceptance either.

The clean solution used by most established freelancers is the signed-proposal hybrid: a proposal whose final page contains the legal terms and an explicit acceptance block ("By signing, Client agrees to the scope, price, and terms above"). One document, one signature, no ambiguity about whether it was accepted. That is a real contract, deliberately, instead of an accidental one. If you are building that final page, start with what should be in a freelance contract.

What to send when: the decision table#

Across the 5 most common solo project scenarios, the right document stack ranges from 1 signed document to 3, and the deciding variables are project size, repeat business, and who is asking.

The pattern behind the table: legal terms should be signed once per client, project definitions once per project, and the proposal exists only for deals you have not won yet. When a repeat client asks for a fresh 6-page proposal for a project you will obviously get, what they actually need is a 1-page SOW.

Freelancers who put agreements in writing are not just safer, they earn more: contract users reported 13.7% higher income than comparable freelancers without contracts in the ILR Review analysis of Freelancers Union survey data. It is correlation, but the mechanism (clear scope, enforceable payment terms) is exactly what this document stack provides.

The reason freelancers skip steps in that table is retyping: the same scope gets written in the proposal, again in the contract, again in the invoice. Tools have mostly solved this. In Raoura, an accepted proposal converts into the contract, the project, or the invoice with one click, so the document stack above is three buttons instead of three retyping sessions (Disclosure: Raoura is our product).

When the law decides for you#

In New York, Illinois, and California, a written agreement is now legally required for freelance work above thresholds of $250 to $800, so "should I send a contract" is no longer always your choice.

Sources: NY DOL, Illinois DOL, California SB 988.

Two details matter for the proposal-vs-contract question specifically. First, the obligation sits on the hiring party, but the practical burden sits on you, because you are the one left unpaid when there is no paper.

Second, a signed hybrid proposal satisfies these laws fine, as long as it contains the statutorily required terms (the work, the price, the payment date). An unsigned proposal plus a "looks great" email is exactly the arrangement these laws were written to end.

The full picture, including what happens when you and the client are in different states, is in freelance payment laws by state.

The signature step: where deals are won and lost#

E-signed agreements close 15% more often and 60% faster than unsigned ones, according to Proposify's 2026 analysis of 742,000 proposals, and e-signatures have carried full legal weight in the US since the year 2000.

The ESIGN Act (15 U.S.C. 7001) says a contract may not be denied legal effect solely because it was signed electronically. There is no legal reason to print, sign, scan, or chase wet ink, and one top-ranking guide still advising "hand-signed for court" is simply out of date. The practical details (what makes an e-signature valid, audit trails, typed vs drawn signatures) are in are e-signatures legally binding.

The same Proposify dataset (State of Proposals 2026) has a less known finding worth stealing: counter-signed proposals, where you sign before sending so the client's signature is the only missing step, closed 65% more often and 25% faster than proposals the sender had not signed. Signing first signals commitment and removes a round trip. It costs you nothing.

Whatever tool you use, the standard to hold it to: the client should be able to read and sign from a phone, both parties should get a copy automatically, and the record should show who signed and when.

Verified July 2026. Contract formation elements from Cornell Law School's Legal Information Institute; close-rate and counter-signing figures from Proposify's State of Proposals 2026 (742,000 proposals analyzed); income figure from the ILR Review analysis of Freelancers Union survey data (Rodgers, Horowitz, and Wuolo, 2014); state law thresholds confirmed against the NY DOL, Illinois DOL, and California Legislative Information; e-signature validity from 15 U.S.C. 7001. The 0-of-4 top-ranking-page observation is our own count, July 2026.