How SaaS teams review redlines against past deals

A customer's lawyer sends your MSA back with edits. Someone on your team says no. Then the account exec says you already gave Acme that exact language last year, and nobody can confirm it either way before the end of the week.

That question has an answer sitting somewhere in your own files. Somebody just has to find it. And if you cannot find it quickly, you either hand over the term to keep the deal moving or you stall the whole thing while somebody goes digging.

Legal teams are absorbing more of this without more people. In the Thomson Reuters Institute's 2024 Legal Department Operations Index reporting, 79% of corporate law departments said matter volumes rose over the prior year, and close to two thirds reported flat or shrinking attorney headcount.

What is actually sitting in a SaaS review queue

Your queue is mostly the same few templates coming back with somebody else's edits on them, over and over.

For a single account, the terms are spread across more files than people expect. There is the MSA. There is an order form for the original purchase, plus another for every expansion and renewal. There is usually a DPA and a security addendum, because the customer's security team demanded them. There may be a service level exhibit. On larger deals there is often a one-page side letter that somebody senior signed to close the quarter.

So when a customer asks for something, the thing you already agreed to is frequently sitting in one of those side files and not in any signed MSA. That changes where you have to look.

Where the playbook and the signed record come apart

A playbook lists positions. It tells a reviewer what to ask for, what to accept, and what to send upstairs. You should have one, and it should be specific enough that software can apply it, which is a separate exercise we walked through in writing playbook entries a machine can follow.

But a playbook drifts. Exceptions approved under end-of-quarter pressure land in a signed file, and the entry often never gets updated to say the company now grants this above a certain deal size. The playbook keeps saying no while your executed record says yes.

Reviewers working only from the playbook can then push back on things the company already conceded. Sales notices. Once that happens, nobody internally takes the playbook seriously, and that is a harder hole to climb out of than the term you were arguing about.

A check you can run on every redline

Before pushing back on a customer's edit, run the proposed language against your executed set: MSAs, order forms, addenda, side letters. You are looking for one of three answers.

Never granted. Hold, and now you can tell the account exec that with something behind it.

Granted, but narrowly. This is the useful case. You gave something close once, capped, or only for one data type, or only above a certain contract value. Offer that version instead of arguing about the customer's.

Granted broadly. The playbook is out of date. Stop defending a position the company walked away from and rewrite the entry.

The fights also repeat. A Commerce and Contract Management Institute study, from the body co-founded by NCMA and World Commerce & Contracting, notes that commercial business-to-business negotiations tend to be dominated by a short set of terms: limitation of liability, indemnities, intellectual property rights and data security. If the same handful of items come back on deal after deal, you have built precedent on them whether anyone was tracking it or not.

Doing the lookup by hand is the slow part. Nobody wants to spend a Tuesday opening three years of countersigned PDFs to answer one question from sales. Software will not tell you whether to concede the term, but it can read the related documents and show you where similar language already sits. The call stays yours.

The number worth watching

Track how often the precedent check changes the answer.

If it rarely does, your playbook matches what you actually sign, and you can save the check for escalations.

If it changes the answer often, the playbook has become a work of fiction and the fix is upstream. Rewrite the entries that keep losing. Stop paying reviewers to defend positions the company gave up on months ago.

Deals will still take as long as they take. What you get back are the days you currently lose to a question your own files already answered.

Run a customer redline through goHeather free and see what it pulls up.

This is legal information, not legal advice; consult a lawyer for legal advice.