Why a Big Builders’ Merchant Couldn’t Stop Its Salesman Joining a Rival: A Lesson for Every Business Owner

If you employ salespeople, account managers, or anyone who holds your customer relationships, you have probably wondered what stops them walking out of the door and straight to a competitor with your customers in tow. The usual answer is the restrictive covenant: the clause buried near the back of the employment contract that says, in effect, “you can’t compete with us for a while after you leave.”

A High Court judgment handed down on 17 June 2026, Huws Gray Ltd v Gentleman, is a salutary reminder that having such a clause is not the same as having one that works. Huws Gray is one of the country’s largest builders’ merchants, with more than 250 branches and around 4,300 staff. It had a non-compete clause. It went to court to enforce it. And it lost, comprehensively, because the clause was drafted far too widely.

For a small business owner, this is not a dry piece of legal trivia. It is a worked example of how the protection you think you have can evaporate at exactly the moment you need it, and why the key is in the drafting.

What happened

Daniel Gentleman was an Area Sales Manager for Huws Gray, responsible for the customer relationships at three of its branches near where he lived, in Swindon, Newbury and Cirencester. By the end of his employment he looked after 103 of the higher-spending customer accounts. He resigned in November 2025 and went to work for MKM Building Supplies, a direct (if smaller) competitor, as the external sales representative for MKM’s new Swindon branch.

His contract contained a six-month non-compete clause. Huws Gray tried to enforce it, and also asked the court to stop him misusing what it said was confidential information, in particular its internal pricing. The judge, HHJ Russen KC, refused on every count. He accepted that Huws Gray had a perfectly legitimate interest in protecting its customer relationships, but held that the clause it had chosen to do that job was an unlawful restraint of trade, and therefore void and unenforceable.

That distinction is key. The employer did not lose because it had nothing worth protecting. It lost because the clause wasn’t drafted as well as it could have been and reached far wider than it ought to have done.

The clause itself

It helps to see what the actual wording looked like, because the lesson is in the drafting. The non-compete said:

“You undertake that during the period of your employment with the Company and for six (6) months from the termination of your employment with the Company (however that comes about) you will not without the prior written consent of the Company, engage or be concerned or interested, whether directly or indirectly, and whether as principal, partner, employee, adviser, agent, consultant or otherwise, in any trade or business that competes or is preparing to compete with any business carried on by the Company or any Group Company … where such Competing Business is located within 20 miles of the Company’s branch or branches for which you had responsibility in the six (6) months prior to the termination of your employment.

You shall be free to engage in any business so far as your duties and work shall relate exclusively to work of a kind which is not related to any area in which the Company has developed Confidential Information, and in which you have not been involved during your employment by the Company.”

The first paragraph is the restriction. The second paragraph, which the parties called “the Carve-Out”, was meant to soften it, by giving the employee back some freedom to work.

Read it slowly, because the judge had to read it “a number of times” himself, and that difficulty is what sank it.

Why the clause was too wide

The court found several fatal problems. Taken together they are a checklist of how not to draft.

It caught any role, anywhere in the competitor. On its wording, the clause stopped Gentleman taking any job with a competing business, not just a sales job near his old patch. It would have stopped him working in the competitor’s HR department, its finance team, its builders’ yard, or its head office in Hull, hundreds of miles away. A clause meant to protect three branches’ worth of local customer relationships was instead trying to wall the employee off from an entire company in every capacity.

It protected parts of the business he had nothing to do with. The clause covered “any business” of Huws Gray, not just builders’ merchanting. The judge noted it would have stopped Gentleman working for a kitchen retailer like Howdens, or even an online retailer such as Amazon or Tesco that happens to sell loft insulation, regardless of whether he had ever touched that side of things.

The carve-out carved out nothing. This is the part business owners should sit up for. The “freedom” the second paragraph appeared to give was an illusion. Because of the way it was written, the employee could only rely on it for work completely unrelated to any area where the company had developed confidential information, and the company’s own manager admitted in evidence that there was no part of the business without confidential information. So the exception could never actually apply. The judge’s verdict was blunt: “The Carve-Out is really nothing of the sort.” A get-out clause that can never be used is worse than none, because it invites a court to conclude the restriction is as suffocating as it reads.

“And” did not mean “or”. The employer tried to argue that the word “and” in the carve-out should be read as “or”, which would have widened the employee’s freedom and might have saved the clause. The judge declined. There is a principle that lets courts lean towards an interpretation that keeps a clause valid, but only where the words can genuinely bear that meaning. As the employee’s barrister put it, if your spouse asks you to buy a black-and-white t-shirt, coming home with one that is only black does not satisfy the request. “And” meant “and”.

Six months from day one, even on a one-week notice period. Reasonableness is judged as at the date the contract was signed, not with hindsight. This clause bit from the employee’s very first day, including during his probation, when he could be dismissed on one week’s notice and had not yet built up any of the customer relationships the clause was supposedly there to protect. A restriction that would be plainly excessive against a brand-new starter who might be gone in a fortnight is vulnerable, because the court tests it against that starting position. (This is sometimes called the “Quilter point”, after an earlier case making the same point about a financial adviser still in her probation period.)

The “we can give consent” wording counted for nothing. The clause said the employee couldn’t compete “without the prior written consent of the Company”. The employer argued this softened the restriction. The judge disagreed: an employer’s completely free, unconstrained power to say yes or no doesn’t make the underlying restriction any narrower. He compared it to a “no trespassing” sign, which is not made any less of a barrier just because the landowner occasionally lets someone cross.

The clause that would have worked, and why it didn’t

Here is the sharpest lesson of all. Gentleman’s contract also contained a different, more targeted set of restrictions, a non-solicitation and non-dealing clause. These are the clauses that say “you may not try to win away, or do business with, our customers” for a period after you leave.

The courts generally treat these far more favourably than blanket non-competes, precisely because they protect the real asset, the customer relationship, without stopping the person earning a living. The judge made the point that a properly drafted version of this clause would very likely have given Huws Gray exactly the protection it actually wanted.

So why didn’t the employer rely on it? Because it was unusable. The clause was built around the terms “Restricted Customer” and “Restricted Potential Customer”, and the contract never defined either term. As the judge put it, the provision was therefore “meaningless”. The employer had the right tool in the box, and it was broken.

Worse, the existence of that broken-but-better clause actively damaged the employer’s case on the non-compete. The judge reasoned that because Huws Gray had clearly contemplated protecting itself with a targeted customer clause, it was poorly placed to argue that a far blunter, wider non-compete was “reasonably necessary”. The narrower tool’s mere presence was evidence that the wider one went too far.

The confidential information claim also failed

Huws Gray also argued Gentleman would misuse confidential pricing, in particular, an internal figure called the “AD Cut-Off”, the lowest price a regional manager could approve without going higher up. The court dismissed this too, describing the case as “baseless”.

The reasoning is instructive for any business that thinks its prices are a protectable secret. The judge accepted the employee’s evidence that day-to-day sales work doesn’t depend on knowing your own employer’s floor price; it depends on whether you can match or beat the rival’s quote in front of you. Customers routinely shop quotes around between merchants and volunteer competitors’ prices to get a better deal. A competitor wants to know the price it has to beat, not the internal mechanics of how you arrived at yours. And a customer’s contact details are not your property: customers typically hold accounts with several suppliers, and their details are publicly findable. The court even rejected a request to make the employee delete his LinkedIn connections, noting he could simply re-add them with a few clicks.

What this means for your business

You do not need to be a 250-branch national chain for these lessons to bite. If anything, they matter more to a smaller business, where losing one salesperson and their accounts can be existential. A few practical takeaways:

• A non-compete is the bluntest instrument, and the hardest to enforce. If what you really want to protect is your customer relationships, the clause that does that job is a well-drafted non-solicitation and non-dealing clause aimed at the customers that person actually dealt with, not a blanket ban on working for a competitor.
• Definitions are not optional. The most damning failure in this case was a clause that fell apart because two key terms were never defined. A contract full of impressive-sounding protections that reference undefined terms protects nobody.
• Wider is not safer. There is a natural instinct to draft restrictions as broadly as possible to “cover everything”. The opposite is true. A court faced with an overreaching clause does not trim it back to something reasonable; it strikes the whole thing out. A narrow clause you can enforce beats a sweeping one you cannot.
• Think about day one, not just your star performers. Because reasonableness is judged from the moment the contract is signed, the same six-month restriction applied identically to a brand-new probationer and a ten-year veteran is hard to defend. Consider tailoring restrictions to seniority or length of service, or only having them start after probation.
• If you have standard employment contracts you have been reusing for years, this case is a good prompt to have the restrictive covenants looked at properly before you next need them, rather than discovering their weaknesses in a courtroom when a key employee has already left.

The full judgment is published on the National Archives’ Find Case Law service and on BAILII: Huws Gray Ltd v Gentleman [2026] EWHC 1309 (Comm).

I’m Steven Mather and I help small and medium sized, owner-managed businesses with their legal issues – like drafting restrictive covenants that work. Get in touch if you think I might be able to help your business.