A company offers £5,000 for a returned prototype, then refuses to pay because nobody signed anything. Here is why they owe it.

A company posts a reward: £5,000 to anyone who returns a prototype device stolen from its facility. Four weeks later someone hands it in. The company, having recovered the device, declines to pay, arguing nobody ever agreed to anything — there was no negotiation, no signature, no exchange of promises.
They are wrong, and the reason is the distinction this article is about. The reward was a unilateral offer. It did not need a promise in return. It needed performance, and performance is what it got.
Most commercial agreements are bilateral, so the unilateral category feels academic until you encounter one — in a reward, a commission structure, an insurance policy, a bonus scheme, or an option. Then the difference determines when the contract formed, who could withdraw and when, and whether anyone owes anything at all.
The difference is in what the offeror is asking for.
The practical test is to ask what acceptance looks like. If the offeror wants to hear "yes, I agree", it is bilateral. If they want the thing done, it is unilateral.
The overwhelming majority of commercial agreements. Employment contracts, supply agreements, leases, sales contracts, service agreements, software subscriptions. Both parties exchange promises, both are bound from formation, and both can sue the other for failing to perform.
Formation follows the standard sequence: offer, acceptance matching the offer, consideration on both sides, and the remaining requirements set out in the six elements of a valid contract. Because both sides are bound simultaneously, either can breach, and remedies flow in both directions.
The mutual commitment is what makes bilateral contracts useful commercially: each party can plan on the strength of the other's promise before anything has been delivered. A manufacturer can order raw materials on the strength of a purchase commitment, because the commitment is itself enforceable.
Less common, but they appear in recognisable forms:
A defining feature. In a bilateral contract, acceptance must generally be communicated to the offeror. In a unilateral contract, performance itself is the acceptance — the person returning the device does not need to have notified the company in advance that they intended to look for it.
This also means multiple people can attempt performance, and the offer is generally accepted by whoever completes first, unless its terms provide otherwise.
The most legally interesting difference, and the one that generates real disputes.
An offer can normally be revoked any time before acceptance. In a unilateral contract, acceptance is completed performance. Taken literally, that means the offeror could withdraw the promise when someone is 95 percent of the way through the act — having watched them do nearly all the work.
Courts have found that outcome unacceptable and have developed limits on it. The prevailing approach in most common law jurisdictions is that once the offeree has begun performance in a meaningful way, the offeror cannot revoke, either because an implied secondary obligation to keep the offer open arises, or because beginning performance is treated as accepting on condition of completion. The offeree, importantly, is still not obliged to finish.
The boundaries are fact-sensitive. Mere preparation to perform generally does not lock the offer; substantial commencement generally does. Where the line sits between those has been litigated repeatedly, which is a reason to draft explicitly rather than rely on the default.
Where a unilateral offer was made publicly — an advertisement or public notice — the offeror generally must use comparable publicity to withdraw it. Notifying one person is not enough, and it is not necessary to prove every potential offeree saw the revocation.
Worth noting that many real agreements resist neat classification. A contract may begin unilateral and become bilateral once performance starts. Long-term agreements frequently contain both structures — a bilateral services agreement with a unilateral bonus provision attached. Some legal systems have moved away from the distinction as an organising principle, treating it as descriptive rather than determinative.
The practical implication is that the label matters less than the drafting. If you want someone bound to perform, say so explicitly and take their promise. If you want to promise a payment on a result without obliging anyone to pursue it, say that instead, and address expressly whether and when you can withdraw. Explicit terms displace the default rules the classification would otherwise supply.
In a bilateral contract both parties exchange promises and both are bound from the moment of agreement. In a unilateral contract one party promises something in exchange for an act, and the other party accepts by performing rather than by promising — so only the offeror is bound until performance occurs. The clearest test is to ask what the offeror wants in return: a promise, or the thing done.
Bilateral. The employer promises to pay and provide the agreed terms; the employee promises to work. Both are bound from the start and either can breach. Specific provisions within an employment relationship may be unilateral in character — a discretionary bonus promised on achieving a target, for instance, where the employee is not obliged to achieve it — but the contract as a whole is bilateral.
Before performance begins, generally yes, provided the revocation is communicated — and for a publicly made offer, communicated with comparable publicity. Once the offeree has substantially begun performance, most common law jurisdictions prevent revocation, on the reasoning that the offeror should not be able to watch someone do the work and then withdraw. Mere preparation usually does not trigger this protection; substantial commencement usually does, though where exactly the line falls is fact-specific.
It can be. A reward is the textbook unilateral offer: a promise of payment for a specified act. If the offer was sufficiently definite, was communicated, and someone performed the act, a contract generally forms and the reward is payable. Whether the performer needed to know about the offer before acting is treated differently across jurisdictions. Vague or plainly promotional statements may be treated as puffery rather than a serious offer, which is a common line of defence.
It is commonly analysed that way. The insurer makes a binding promise to pay on the occurrence of a defined event, while the insured is generally not obliged to continue paying premiums — they simply lose cover if they stop. Some commentators and some jurisdictions treat policies as bilateral, particularly where the insured undertakes positive obligations such as disclosure, notification, and risk mitigation. The classification is rarely decisive in practice, because policy terms are express and detailed.
Less than it once did as an organising principle, but it still determines real outcomes — particularly around when a contract forms, whether an offer can be withdrawn, and whether the performing party has any obligation at all. Some jurisdictions have de-emphasised the categories in favour of analysing offer, acceptance, and consideration directly. The reliable approach for commercial drafting is not to rely on the classification but to state expressly what each party must do, when the arrangement becomes binding, and on what terms it can be withdrawn.
HERO is built for agreements where the structure carries meaning — where which obligations are conditional, which are mutual, and which survive a change elsewhere in the document should be visible rather than buried in prose. Book a demo.