Somewhere in most employment contracts, usually near the end where nobody reads carefully, sits a clause promising that the employee won't compete with the business, poach its clients, or hire away its staff for some period after they leave. It feels like protection. Whether it actually is depends entirely on how it's written — and a huge number of these clauses, in practice, aren't worth the paper they're printed on.
What a restrictive covenant is for
Restrictive covenants exist to protect a small number of genuine business interests: confidential information and trade secrets, client and supplier relationships the business has invested time and money building, and the stability of a trained team. They are not meant to stop a former employee earning a living in their chosen field — and any covenant that tries to do that, rather than protect one of those specific interests, is exactly the kind a court is likely to strike down.
Why UK courts start from suspicion, not support
Restrictive covenants sit awkwardly against a basic legal principle: contract terms that restrain someone's ability to trade are, by default, unenforceable, unless the business can justify them as reasonable and no wider than necessary to protect a legitimate business interest. That means the burden is on the employer to prove the clause is fair, not on the ex-employee to prove it's unfair. A covenant that's too broad in what it bans, too long in duration, or too wide geographically will typically be cut down or thrown out entirely — and a court won't usually rewrite an unreasonable clause into a reasonable one for you; it just refuses to enforce it.
A non-compete clause that would stop someone doing almost any job in the industry, almost anywhere, for years, isn't a strong protection. It's a clause a tribunal will read once and set aside.
The four common types, and how far each usually stretches
Non-compete clauses (banning work for a competitor) are the hardest to enforce and courts scrutinise them most heavily, because they most directly restrict someone's livelihood — they tend to only survive when narrowly drawn around a senior role with real access to sensitive strategy. Non-solicitation of clients (banning approaching clients the employee dealt with) is more commonly enforceable, provided it's limited to clients the person actually had contact with, not the entire client book. Non-solicitation of staff (banning poaching colleagues) tends to be viewed similarly. Confidentiality clauses (protecting specific trade secrets and client information) are usually the most robust of the four, because they protect information rather than restrict someone's ability to work at all.
What actually makes a covenant hold up
The covenants that survive a challenge tend to share three features: they're limited to a genuine, identifiable business interest rather than general competition; they're proportionate in scope, geography and — critically — duration, with three to twelve months being far more defensible than two or three years for most roles; and they're tailored to the specific person's actual role, rather than a single boilerplate clause bolted onto every contract regardless of seniority. A junior team member with no client contact and a departing sales director should never be bound by the same non-compete wording, and a court will notice if they are.
What to do if you're actually worried about someone leaving
If a valued employee is leaving and you're worried about client poaching, the covenant in their contract is your starting position, not your only tool. A firm but calm reminder in writing of what the contract says, sent promptly rather than months later, carries real weight even before any legal step is taken — plenty of disputes never go further than that letter. Reviewing what confidential information the person actually had access to, and whether it's been protected properly (password changes, access revoked on their last day), matters as much as the legal wording. And getting early advice on whether your specific covenant is actually enforceable — before threatening legal action you can't actually back up — avoids the credibility-damaging outcome of making a threat that gets laughed out of a solicitor's letter.
The lesson for the next contract you write
The best time to fix a weak restrictive covenant is before you need it, not during the dispute where you discover it doesn't hold up. Getting contracts reviewed and tailored to each role, rather than reused from a template found online, is one of the more valuable pieces of routine legal spend a growing business can make — see also our guide on hiring your first employee for what else that first contract should actually cover.
Garden leave as the underused alternative
Where a departing employee genuinely has access to sensitive, time-limited information — a live pitch, an upcoming product launch, negotiations with a key client — garden leave is often a more reliable protection than a restrictive covenant, and it's underused by smaller employers who don't realise it's an option. Putting someone on garden leave during their notice period, on full pay but away from clients, systems and colleagues, buys time for sensitive information to age out of relevance without relying on a covenant's enforceability being tested after the fact. It only works if the contract allows for it, so it's worth building the clause in before you need it, alongside the restrictive covenants themselves.
What this means for the contract you write next
None of this is an argument against having restrictive covenants at all — a well-drafted, proportionate one is a genuinely useful deterrent even when it's never tested in court, because most departing employees would rather not risk a legal dispute over client poaching if they can help it. The argument is against copying a generic, sweeping clause into every contract and assuming it does the job. A short, specific non-solicitation clause tied to real client contact, reviewed when someone's role changes significantly, does more real-world protecting than an intimidating-sounding non-compete that would collapse the moment anyone actually challenged it.
