Non-competes, non-solicitation and garden leave — why role changes expose gaps in old drafting, and what to check before a key person leaves.
A key employee resigns on a Monday morning. They know your pricing, your best clients and half your team by first name. Somewhere in a contract they signed years ago is a clause meant for exactly this moment. The question business owners ask us is the same: does it actually stop them? The honest answer is that it depends far more on how that clause was written, and when, than most people expect.
Restrictive covenants are the clauses that survive the end of employment: the promises not to compete, not to poach your clients, not to take your staff. Drafted well, they buy the breathing space to protect what matters while a leaver’s knowledge goes stale. Drafted badly, or never reviewed as a role grew, they can be worth very little at the exact moment you need them. And the courts start from suspicion: a post-termination covenant is void as an unreasonable restraint of trade unless you can show it protects a legitimate business interest and goes no wider than necessary to do so. The burden sits on you, the employer.
Covenants tend to fail for a handful of recurring reasons. Here are the five that come up most, and what to do when someone actually resigns.
1. The restriction doesn’t match the business risk
There are four main covenant types, and they do different jobs. A non-compete restricts carrying on, or being materially involved in, specified competing activities for a limited period (and usually needs a carve-out so it doesn’t accidentally catch a small passive shareholding in a listed rival). A non-solicitation stops the leaver soliciting or enticing away customers they had real dealings with. A non-dealing goes wider and stops them dealing with those customers even where the customer approaches them. And a staff non-solicitation covenant stops them soliciting or enticing away your genuinely key people, where workforce stability is a real interest to protect.
The instinct is to reach for the non-compete because it feels strongest. In practice it is the widest and the hardest to enforce, because it stops someone earning a living in their field. The narrower covenants are often the ones that hold. Match the covenant to the interest you are actually protecting: if the risk is your client list walking out the door, a well-drafted non-solicit and non-dealing may protect you far better than a broad non-compete a court will not uphold. And a covenant can only protect a genuine interest — your trade connections, your confidential information and trade secrets, or the stability of your workforce. It cannot shield you from ordinary competition, or stop someone using the general skill and experience they built up while working for you.
2. It was unreasonable when it was first agreed
Reasonableness is judged on scope, duration and geography, and — this is the point that catches people out — as at the date the covenant was entered into, not the date you try to enforce it. That assessment is forward-looking: a court can take account of the role and the progression the parties could reasonably have contemplated when they signed. But a covenant that was too wide for the role when it was imposed is not rescued because the person later becomes senior enough to justify it.
Duration is measured against how long the connection or the information actually stays live: a client relationship that turns over every few months does not justify a two-year restraint. Geography matters for some businesses and very little for others — for an online or national business the sharper questions are usually what activities are restricted, which customers and staff are covered, and over what look-back period. A nationwide restriction for a genuinely local business will need particularly strong justification and may well be wider than necessary.
|
Wider is not safer. An over-reaching covenant is more likely to be struck out in full, leaving you with nothing, than a tighter one a court will actually enforce. If the offending words can be deleted without rewriting the clause or materially changing the overall restraint, a court can sometimes remove them and enforce the rest — but it will not add words to save it, and it will not rewrite the bargain or salvage a fundamentally overreaching package. Never draft on the assumption that a judge will fix an overreaching covenant. |
3. A promotion or contract change was mishandled
Here is the correction to a common myth: a promotion does not automatically kill a covenant. A valid restriction does not fail simply because the person becomes more senior, and — as above — an invalid one is not cured by it either. What a promotion does is expose whether the covenant was properly drafted in the first place, and whether it still fits. The restricted business, the customer definitions or the protected interests may no longer match the role, or a later contract may have quietly superseded the terms you were relying on.
A promotion, then, is a review trigger, not an expiry event. And if you do refresh covenants, mind the consideration point: a new or widened restriction needs a genuine new benefit given in exchange for it. Don’t assume continued employment, or an ordinary annual pay review the person was getting anyway, will do the work. A promotion, a specific pay rise, a bonus or an option grant can — but only if it is expressly given in return for accepting the new covenants, and the person signs up before it takes effect. Slip the new terms in afterwards and you may have covenants with nothing behind them.
|
In practice (illustrative). Two situations, two outcomes. A junior signs a twelve-month worldwide non-compete covering the whole group; it was excessive for the role then, and five years of promotions do not make it reasonable now. By contrast, a proportionate customer non-dealing restriction, tied to clients the person actually dealt with, may keep working as they rise — but it may say nothing about the confidential strategy, key staff or new divisions they now control. Either way, promotion is the moment to look, not to assume. |
4. Garden leave and the exit mechanics don’t work together
Garden leave — keeping someone employed and paid, but away from the business during their notice — is a genuine protection, provided your contract actually gives you the power to impose it. Their client contact stops and their knowledge goes stale while they are still bound by the duty of fidelity and their confidentiality obligations. But it interacts with the post-termination covenant, and there is a myth to clear up here too: there is no automatic legal rule deducting garden leave from the restraint that follows. Some contracts expressly reduce the covenant by time spent on garden leave; without that wording, the periods may run consecutively, though a court can still weigh the overall restraint when deciding what is reasonable. Design the two together rather than bolting them on separately.
The bigger exit-mechanics trap is your own conduct. If you commit a serious (repudiatory) breach of the contract — mishandling notice or pay in lieu, withholding money that is due, a botched summary dismissal — and the employee accepts it, you can lose the right to enforce otherwise valid covenants altogether. It is exactly when an owner is angriest, having just learned the person is off to a competitor, that this damage gets done. Check the contract’s powers and payment obligations before you act.
5. You can’t prove what you’re protecting — or what they took
Covenants are one layer, not the whole wall. Running alongside them should be confidentiality obligations that survive termination, and — where someone has created code, designs or content the business treats as its own — a proper handle on intellectual property. The rules differ by who created the work: for copyright made by an employee in the course of employment, the employer is usually the first owner automatically, subject to any agreement otherwise; for a contractor or agency, paying the invoice does not transfer copyright, and you generally need a signed written assignment. Our Contracts and IP article covers the ownership chain in full.
And keep evidence. If it ever comes to protecting your position, you will need to show what the person had access to, what left the building, and why the information still has value. That is far easier to assemble in the first hours than to reconstruct months later.
What to do the day you find out
Find the actual signed contract — not the template you think they signed — and check what covenants exist, whether they were ever reviewed, and how long they run. Preserve the evidence before anyone touches systems or wipes a device: relevant email, downloads, access and forwarding logs. Secure access to client data, pricing and live deals, proportionately. Map the real risk: who they have recently dealt with, what they have seen, how long it stays valuable. Check the termination mechanics before you make a move, so you don’t hand the leaver a way out. And take advice early, while the position can still be shaped.
If this tips into a genuine dispute — an urgent injunction to stop an unfair head start built on your confidential information, for instance — that is contentious work. Our core is the preventive and advisory side: getting the covenants right before anyone leaves, and protecting the position in those first hours. Where court proceedings are needed, we coordinate the right disputes specialists so the employment, corporate and evidence strategy stays aligned.
Quick answers
When is a restrictive covenant enforceable?
When it protects a legitimate business interest — trade connections, confidential information, or workforce stability — and goes no wider than reasonably necessary in scope, duration and, where relevant, geography. Reasonableness is judged as at the date it was signed, and the burden is on the employer.
Does a promotion invalidate an existing covenant?
No. A valid covenant does not fail just because the person becomes more senior, and an invalid one is not cured by promotion either. But promotion is the moment to check whether the wording still fits the role, and whether it needs re-agreeing — with fresh consideration.
Is there currently a statutory cap on UK non-competes?
No. The Government published a working paper on reform in November 2025, which closed for responses in February 2026, but nothing has been enacted and no timetable has been set. Common-law reasonableness still governs. It is worth watching, not planning around yet.
Download the full guide: Restrictive Covenants — Protecting the Business When Key People Leave (PDF), including the covenant review checklist.
This article provides general information on the law of England and Wales and is not legal advice. Law stated as at 29 July 2026.
