Two founders start a limited company, split the shares fifty-fifty, and agree everything else will 'sort itself out' as they go. It usually does — right up until one of them wants to leave, or disagrees with the other about the business's direction, or stops pulling their weight, and there's no document anywhere that says what happens next. A shareholders' agreement is the document that answers those questions while everyone's still getting on, so nobody has to negotiate them for the first time in the middle of a falling-out.
What it actually is
A shareholders' agreement is a private contract between the shareholders of a company, sitting alongside the company's articles of association at Companies House. Where the articles are a fairly generic legal framework, the shareholders' agreement is where the specific, honest arrangement between the actual people involved gets written down — who decides what, what happens if someone wants out, and what happens if things go wrong. It isn't legally required to start a company, which is exactly why so many founders skip it, and exactly why so many later wish they hadn't.
Decision rights: who actually has to agree
The most practical clause in most agreements is a list of decisions that need unanimous or majority shareholder agreement, rather than being left to whoever happens to be running day-to-day operations. Taking on debt above a certain amount, issuing new shares, changing the company's core business, or paying out dividends are common ones. Without this, a majority shareholder can technically make sweeping decisions alone — which might be exactly what you intend, or might be the opposite of what a minority co-founder assumed when they signed up.
Leaver provisions: the clause everyone skips and everyone needs
This is the one that saves relationships. What happens to someone's shares if they leave — whether by choice, by being pushed out, or through illness or death? Good leaver and bad leaver clauses set different terms depending on the circumstances: someone who leaves on good terms after years of contribution is usually treated more generously than someone who leaves early or is removed for cause. Without this written down, a departing co-founder can simply keep their shares indefinitely, meaning the people still doing the work are permanently sharing profit and decision-making with someone no longer involved.
The people who most need a shareholders' agreement are the ones who are certain they'll never need it. That certainty is exactly the moment to get it in writing — not after the friendship has already started to strain.
Vesting: the clause most UK founders have never heard of
Common in venture-backed businesses and increasingly sensible for any multi-founder company: shares are earned over time — commonly across three or four years — rather than handed over in full on day one. If a co-founder walks away after six months, vesting means they leave with a fair slice of what they actually contributed, not an equal permanent stake earned in a fraction of the time everyone else stays. It feels unnecessary when everyone's excited and committed. It's precisely the clause that protects everyone if that changes.
Deadlock: what happens when two equal partners disagree
A fifty-fifty split feels fair at the start and can become a genuine problem the moment the two shareholders disagree on something significant, because neither side can outvote the other. A deadlock clause sets out what happens next — anything from a casting vote for a specific role, to mediation, to a mechanism for one side to buy the other out at a fair valuation. Deciding this in advance, while both sides are reasonable, is far easier than negotiating it in the middle of an actual disagreement.
Drag-along and tag-along: the clauses that matter when you sell
These two rarely get discussed early on, but they decide how a future sale actually plays out. A drag-along clause lets a majority of shareholders force a minority to sell on the same terms, if the majority agrees to sell the whole company — without it, a single small shareholder can technically block a sale everyone else wants, simply by refusing to sign. A tag-along clause protects the reverse case: if a majority shareholder sells their stake, a minority shareholder can insist on selling theirs on the same terms, rather than being left holding shares in a company under new ownership they never agreed to. Neither clause matters at all until the day someone actually wants to sell — at which point, whichever one is missing becomes the single most expensive omission in the whole document.
Picture two founders, sixty-forty, three years in. The sixty-percent founder gets an acquisition offer for the whole business. The forty-percent founder isn't ready to sell, doesn't want to work for the acquirer, and — with no drag-along clause in place — is legally entitled to simply say no, and hold the shares regardless of what the majority wants. The deal doesn't have to collapse, but it now involves solicitors, a renegotiation, and months of delay a five-minute clause agreed three years earlier would have avoided entirely. Nobody in that founding conversation was being difficult or short-sighted; they simply never got round to the document, on the reasonable assumption they'd sort it out if it ever came up. It came up.
What happens without one
Without a shareholders' agreement, disputes default to the company's articles of association and general company law — which are built for the general case, not your specific business or relationship. That usually means slower, more expensive, more adversarial resolution, often via solicitors and sometimes via court, at exactly the point when the business itself needs steady leadership most. It's the legal equivalent of not having anything in writing with your co-founder — a topic our piece on the agreement we never wrote covers from the other side of the same lesson.
Getting it done properly
A shareholders' agreement isn't a DIY template job for anything beyond the simplest structure — get a solicitor to draft or at least review one, because the clauses that matter most are precisely the ones that are easy to get subtly wrong. Expect a proper first agreement between two or three founders to be a modest, fixed-fee piece of work rather than an open-ended one, and treat it as cheap relative to what it's protecting — a small cost early on, against a very expensive one to sort out later. The best time to have this document is when nobody thinks they need it yet. That's also, unhelpfully, the exact moment it's easiest to skip.
