The Deal Room No. 3: Learning How to Pick Your Battles

This series is written for non-lawyers. It is general education, not legal advice, and it does not replace a lawyer. If you have a question about a specific transaction, get a legal opinion.

Pass three is where you stop reading and start changing things. It is also where most clients defer to their attorneys and expect the attorney to make decisions for them.

But that is not your lawyer’s job. Your lawyer is not going to decide this for you, and a good one will not try.

There is a widely held belief that legal exists to say no. I would challenge that. A lawyer’s job is to tell you, in plain terms, what the risk is, what that risk would cost you, and what alternatives exist if any do. That is the job. Whether to take the risk is your decision, and it stays yours however uncomfortable that gets, because it is your business that carries it.

I have worked in-house and as outside counsel, for large companies and small ones, and the same thing holds everywhere. No two businesses share the same interests, the same ethics, or the same leverage. Every deal is different no matter how hard you have polished the template.

There is also a structural reason the decision has to sit with you. An attorney can fire a client, so long as doing so does not prejudice the client in the matter. You cannot fire a risk you have already accepted. Counsel has an exit and the business does not, and that asymmetry is exactly why the person who lives with the consequence is the person who chooses.

Two questions for everything you flagged

Pass two left you with a marked-up document and two piles of notes. So which of it actually matters?

This is where people freeze, because ranking legal risk is not something anyone can do without having watched a lot of deals go wrong.

So do not rank legal risk. Rank foreseeable consequences.

Take each provision you flagged and ask two questions. What does this do if the relationship goes well? What does this do if it goes badly?

The answers worth writing down are business answers, because those are the ones you are qualified to give and the ones nobody can give on your behalf. I lose the engine I need for the next project. I do not see money until a date my payroll cannot survive. I cannot ship on the platform where my audience actually is. My name is personally on this.

Most provisions do nothing at all in the good case. That is what they are for.

That list is what you bring to your lawyer. You supply what you can and cannot live with. They supply how likely it is, what it would cost, and what can be done about it. Neither of you can do the other half, and a deal goes wrong most often when somebody tries.

Vague is a dial. Ambiguous is a fork.

So you know what you cannot live with. What makes a change safe to make?

Start with the distinction that does the most work in this phase, and that non-lawyers tend to run together.

Vague means the boundary is fuzzy while everyone agrees what is being discussed. Reasonable notice. Promptly. Materially. Vagueness is a dial, and turning it is sometimes the only honest way to write a clause about a future nobody can specify. The efforts clauses from Part 2 live here.

Ambiguous means the words support two different readings, and each party can walk away believing a different one. That is a fork in the road, with the parties on separate roads and neither of them aware of it.

There are times when being vague is necessary or preferable. There is no time when ambiguity is acceptable.

The reason is what happens next. Vagueness gets resolved by a standard. Ambiguity gets resolved by a judge. One is a negotiating position and the other is a lawsuit waiting for something to trigger it.

If that sounds familiar, it is the same failure as a bad defined term from Part 1. Both are a failure to say what is inside and what is outside.

Watch for this when you fix things, because the confusion runs in an expensive direction. People try to cure vagueness by adding specificity, and create ambiguity by writing a second provision that does not line up with the first. You had a fuzzy boundary. Now you have two boundaries in different places.

One word, and what it costs to move it

Minimum changes with maximum impact is achievable. It is also where you can do the most damage without noticing, so here is the discipline, shown end to end.

Say the deliverables clause reads:

Developer will deliver Versions for console, mobile and PC.

That is three builds. If you scoped and priced one, you have just found the most expensive sentence in the agreement, and it took eight words to say it.

So change one word. Turn the “and” into an “or.”

Developer will deliver Versions for console, mobile or PC.

Both of those sentences are perfectly coherent. Both describe a real deal that real people sign. One requires three builds and the other requires one, and the whole difference is a conjunction sitting in the middle of a list.

Now read your own revision the way Part 2 taught you to read theirs, because you are not finished and you may have just made this worse.

Which platform? The sentence does not say, and it does not say who chooses. You are reading it as your call. The publisher is reading it as theirs. That is not vagueness, which you could live with. Those are two different readings of the same words, which is the fork this piece just told you never to accept, and you introduced it yourself while fixing something else.

So finish the job. Either name it:

Developer will deliver a Version for PC.

Or keep the optionality and say whose it is:

Developer will deliver a Version for one of console, mobile or PC, as Developer elects by written notice within thirty (30) days of the Effective Date.

That is the shape of a good revision. You did not rewrite the clause, you did not challenge the deal, and you did not hand the other side a paragraph of new language to argue with. You moved one word, noticed what the move opened up, and closed it.

Now find everything that word touched

So you are done? Not close.

A change that is correct in one place and contradicted in another is worse than no change at all, because now the agreement says two things and somebody else will decide which one governs.

So run the sweep. Four steps:

1.  List every defined term in the provision you edited.

2.  Search the document for each of those terms and read every place it appears.

3.  Search for every cross-reference pointing at the section you edited.

4.  In each of those places, ask whether what you just wrote still holds.

On the deliverables example that sweep reaches a lot further than you would guess. Is “Version” a defined term, and does the definition assume more than one of them? Does the Milestone Schedule still list builds for three platforms? Does the fee assume three builds, and if so, are you now doing a third of the work for all of the money, or all of the work for a third of it? Does the definition of revenue compute royalties across platforms that are no longer in the deal? Does an exclusivity provision assume you are shipping on console? Does the marketing commitment promise launches you are no longer doing?

And the one that undoes everything: is there another provision somewhere else that independently requires all three platforms, quietly overriding what you just wrote?

If there is, you did not win anything. You added a sentence that a different sentence contradicts, introducing an ambiguity that creates unnecessary litigation risk.

What your markup looks like when it arrives

Everything so far has been about getting the change right. What about getting it accepted?

Picture your redline landing on the other side. Two hundred changes, and a hundred and ninety of them are preferences about how a perfectly reasonable document phrases things.

The lawyer opening that file does not sort your ten real issues from your hundred and ninety cosmetic ones. They get defensive, then combative, then a little snarky, and now every one of your ten substantive points has to be won through all of that.

You had ten issues and you buried them.

Keep revisions surgical, substantive, and written so they do not challenge the spirit of the agreement. That last part carries an exception worth naming: if the problem is that the parties disagree about what the spirit of the agreement is, you have a much bigger conversation ahead of you, and a redline is the wrong place to have it.

It is worth saying plainly that redlining is not a one-person job. It is collaborative work between you, the other side, and your own business team. People say “opposing party” out of habit. The best deals get negotiated by people who are not behaving like opponents, and treating a markup as combat produces worse paper and slower closings for everyone.

How to send it

Four rules, and they matter more than people expect:

•       Use tracked changes. If you want to be thorough, incremental, cumulative, and clean. Sending back a clean rewrite alone hides what you did and could read as bad faith even when it is not.

•       Make a change where you have a proposal. Leave a comment where you have a question. Do not redline a question, because a change is an assertion and a question is not.

•       Write a short cover note grouped by issue rather than by page number. Three issues explained in a paragraph each will get you further than forty changes explained nowhere.

•       Leave formatting, house style and word preferences alone outside of meaningful housekeeping at the end of the negotiation. You do not win points for style, so focus on the things that actually make the contract easier to read.

When you cannot get the change

Ask for the fallback before you concede the point. A narrower version of what you wanted is a genuinely different request from the one they already turned down, and it lands more often than you would think.

If the answer is still no, you have arrived where this piece started. It is a decision, it is yours, and you should make it on purpose rather than by running out of energy.

Then do the thing almost nobody does. Write down that you took it. A short internal note saying what the risk is, why you accepted it, and what you will do if it materializes. Remember that you cannot walk away from a risk you have committed to, and that note is what lets you plan around it later instead of rediscovering it in the middle of a dispute.

And keep hold of the option people forget they have. Sometimes the answer is that you do not sign.

What you can do now

Read it three times.

Pass one gave you the shape: what kind of deal this is, and where the drafter put the money and the risk. Pass two gave you the substance, one provision at a time, read by someone who stopped assuming the other side meant what you would have meant. Pass three is judgment, which is the part that keeps getting better for as long as you keep doing it.

None of that makes you your own lawyer, and it was never supposed to. It makes you someone who arrives at the conversation knowing what the document says, what it does, and which parts of it you cannot live with. That is a much shorter and much cheaper conversation than the one that starts with “can you read this for me.”

Chapter 2 leaves the anatomy behind and starts on specific deals, beginning with collaboration agreements. They are worth going first, because they are the ones most often done on trust, a shared vision and an email thread.