↓ Download as PDF

A supply agreement comes back from legal and it is thorough. Liability is capped. Indemnities are bounded. The confidentiality clause is immaculate and the governing-law provision has been argued down to something reasonable. It took four months.

Eighteen months later the supplier misses a ramp, or moves the build to a second site without telling anyone, or quietly stops holding the safety stock everybody assumed was there. You go to the agreement looking for the clause that helps. There is not one. There is a clause about something adjacent, drafted beautifully.

Nobody did anything wrong. That is what makes it worth writing about.

There was nothing to brief

Counsel was handed a template and a list of clauses. That is what there was to hand them, because the deal itself had not been settled.

The master agreement is usually the first thing negotiated. It should be close to the last. It gets pulled forward because it is the part that looks like progress — a document, a version number, a redline going back and forth, movement everyone in the room can see. The four things that actually decide whether the relationship works produce no artefact at all for months.

Whether the two business models can coexist, or whether one side needs exclusivity that the other side’s whole model is built on refusing. Whether what you are buying can be defined tightly enough that a third party could quote against it and test to it. Who owns what, agreed before any work has been done that is worth arguing over. And whether both sides want the same shape of relationship in three years — the same volumes, the same margins, the same ownership.

Settle those four and the agreement writes quickly, because it is recording decisions rather than making them. Draft first and the document becomes the place those arguments happen instead: slowly, in writing, through intermediaries, at the worst ratio of cost to progress available to you.

So what reaches counsel is the artefact that does exist. The last agreement, marked up. A clause checklist. A calendar invitation that says can you take a look at this.

And good counsel is expensive

There is a second reason, and it is the more honest one. Good counsel costs money by the hour. Every hour spent explaining the business is billed before anybody has looked at a word of the agreement, and to whoever is watching the invoice that does not look like progress. So the brief gets trimmed to what seems essential: here is the document, here are the three clauses we are nervous about, can you turn it round quickly.

It is a false economy of the ordinary kind. The hour not spent on background comes back as redline cycles, as clauses argued over that were never going to matter, and as the four months. You saved an hour of context and spent a quarter on markup.

And what you get is law. Against that brief a good lawyer will protect you against generic risk, because generic risk is the only kind they have been shown — accurate, defensible, aimed at nothing in particular. That is not a failure of skill; it is the only responsible thing to do with a brief that contains no business. Good counsel is a different product. It is available only to someone who has been told what the business is trying to do, and you decided that part was too expensive to buy.

What good counsel is actually asking for

In my experience good counsel is openly frustrated by pro-forma work and will say so, given the opening. What they ask for is not more legal input. It is the business: what the deal is for, where the money is, which risks you are being paid to take, and what you would walk away from.

A liability cap is a legal question with no legal answer. Whether the number on the page is enough depends on what one line-down event costs you, and that is an operations figure you already have. Exclusivity is worth fighting for or worth conceding depending on whether the supplier’s business model can accommodate it at all — a commercial judgment you have almost certainly made and almost certainly not written down. And the length of a termination notice period is a function of how long it takes to qualify an alternate source.

Ninety days is the number that appears in most templates, on both sides, largely because ninety days is the number that appears in most templates. If qualifying a second source for that part takes nine months — a new tool, a first article, a customer requalification that is not in your gift — then ninety days’ notice is not an exit. It is a cliff with a date on it. The fix is one sentence and it costs nothing at signature: notice no shorter than qualification. Nobody writes that sentence unless somebody has told the drafter how long qualification takes.

The item that surprises people is the fourth one: telling a lawyer which risks you are content to carry. It sounds like disarming yourself and it is the opposite. It is the instruction that lets them stop. A brief listing only what you fear reads as a demand for total protection — and total protection is slow, expensive and unwinnable. It is also why the redline came back with twenty issues in it, every one marked material and none of them ranked.

What changes

Three things, and the first is counter-intuitive. The redline gets shorter. A lawyer who knows which three risks matter stops defending against the other seventeen, and the document stops growing in the places that were never going to bite you.

The second is that the argument moves to where it belongs. Instead of a business disagreement being conducted at one remove — through markup, over four months, at professional rates — it happens in a room, between the people who can actually settle it, before anyone drafts. The agreement then records the settlement, which is the job it is genuinely good at.

The third is the one worth paying for. Counsel who understands the deal will tell you which of your commercial asks is not going to survive contact with this supplier, and which one you are underplaying. That is advice about the negotiation rather than about the document, it is available from nobody else in the process, and only ever to a lawyer who has been told what you are trying to achieve.

The caveat, since I arrive with a checklist

My own agreement review runs to sixty-six items across nine sections, and some version of it turns up in every engagement. Used as sixty-six boxes to tick it will produce precisely the agreement described at the top of this piece: thorough, late, and protecting the wrong things.

A checklist tells you where money and risk usually hide. It cannot tell you which of them matter in this deal. That judgment is the brief, and the brief is yours. The items you consciously decide not to pursue are as much a part of the review as the ones you chase — write down why you dropped them, because the next person will need the reasoning and will otherwise give the clause away for nothing.

The contract is not the control system. The control system is the specification, the qualification gate, the data you get every week and the relationship you run. But you find out whether the agreement was any good on the worst day of that relationship — and by then the only thing that helps is whether somebody, once, sat down and told the lawyer what the deal was.

That is the work I do.