This guide shows you how to build one from scratch, what to put in it, and how to keep it from rotting.
Why a clause library beats a folder of past contracts
"Find a similar contract and copy the clause" is how most teams operate. It fails in three ways: the precedent you find may itself be a bad compromise, you inherit its drafting errors, and you have no record of why the clause says what it says. A clause library fixes all three by capturing the team's intended position once, with its reasoning, and letting everyone draw from it.
The payoff compounds. Every negotiation that ends in an agreed compromise should feed a new fallback into the library, so the team gets smarter with each deal rather than starting cold.
The anatomy of a clause-library entry
Each entry should carry more than the words. A useful entry has six parts:
- Preferred position — the clause exactly as you would draft it if the counterparty agreed to everything.
- Acceptable fallbacks — ranked, from best to worst, with the precise language for each.
- Walk-away line — the position below which you do not sign, and who must approve any exception.
- Rationale — why each position matters, in one or two sentences a non-lawyer can understand.
- Negotiation notes — common counterparty arguments and how to answer them.
- Trigger conditions — when this clause applies (e.g. only above a deal value, only for data-processing engagements).
The rationale field is the one teams skip and the one that matters most. A fallback without a reason is a number nobody can defend under pressure.
The clauses worth standardising first
You cannot library everything at once. Start with the clauses that are negotiated most often and carry the most risk. For most commercial teams that is:
- Limitation of liability — cap level, mutuality, and the carve-outs.
- Indemnities — scope, mutuality, and whether they sit inside or outside the cap.
- Termination — for convenience, for cause, notice periods, and what survives.
- Intellectual property — background vs foreground IP, licence scope, ownership of deliverables.
- Confidentiality — definition, exclusions, and duration after termination.
- Data protection — controller/processor roles, sub-processing, and transfer mechanisms.
- Payment — terms, late-payment interest, and price-increase mechanics.
- Governing law and dispute resolution — forum, arbitration vs litigation, and seat.
These eight cover the large majority of redlines on a typical commercial agreement.
Worked example: limitation of liability
Preferred position. Liability of each party capped at the total fees paid in the 12 months preceding the claim; consequential, indirect and loss-of-profit damages excluded; cap and exclusions mutual.
Fallback 1. Cap raised to 150% of annual fees. Acceptable — the exposure is still proportionate to the contract value.
Fallback 2. Data-breach liability carved out of the cap but subject to a separate super-cap (e.g. 3x fees). Acceptable where the engagement involves limited personal data.
Walk-away. Uncapped liability of any kind, or indemnities sitting entirely outside the cap with no super-cap. Requires General Counsel sign-off to accept.
Rationale. The cap keeps worst-case exposure tied to the revenue the deal actually generates. An uncapped indemnity can dwarf the contract value many times over and is uninsurable in practice.
Negotiation note. Counterparties often argue "we need uncapped liability for breach of confidentiality." Counter: agree a confidentiality super-cap rather than removing the cap entirely — it addresses their concern without exposing you to unlimited risk.
Governance: keeping the library alive
A clause library that is not maintained becomes a liability — people trust stale language. Three rules keep it honest:
- One owner per clause. Someone is accountable for each entry being current.
- A post-deal feedback loop. Every negotiated compromise is reviewed: does it become a new fallback, or was it a one-off exception?
- A quarterly review against the law. Statutory caps, data-transfer mechanisms and consumer-protection rules change. Diary a review so the library never recommends something now unenforceable.
From playbook to live review
A playbook is only as good as your team's discipline in applying it. The hard part is consistency — making sure every contract that crosses the desk is actually checked against the agreed positions, not just the ones the reviewer remembers.
Attorly turns your playbook into an automated reviewer. Load your preferred positions and fallbacks, and it checks each incoming contract against them, flags where the counterparty's language falls below your walk-away line, and drafts the redline that moves it back toward your preferred position. The same playbook applies across 13 jurisdictions across Europe, the Nordics, the UK and the US, so a team operating in multiple markets enforces one consistent standard.
Build your playbook once, apply it everywhere, at attorly.ai.