River
Y CombinatorBacked by Y Combinator
FREE TEMPLATE

Contract Playbook and Position Tracker

Five documents and four sheets that build every position out of the contracts you already signed, and print the count that makes each tier true.

Free download  ·  No account needed

Business Team Guidance

Enterprise order form  ·  what you can agree without asking

Derived from [n] executed contracts. Every line carries the count, because the count is the part you can say out loud.

If they ask forAnswerThe count behind it
[the position the register calls acceptable]Yes, on the call[n] of the last [n] closed here
[the position the register calls preferred]Yes, obviously[n], never escalated
[a term below the walk-away]No. Goes to [named approver][n] crossings, each approved
[a term the playbook has no row for]Ask. There is no position yet[n] instances, too thin to read

Why the number is on the page

A salesperson can say the count out loud in a negotiation. Our policy is [the preferred position] is not a sentence that survives contact with a counterparty who has already signed one of the deals that went the other way.

Every bracket fills from your own executed history. Nothing here ships with a position already in it.

A playbook ships three tiers for every clause: preferred, acceptable, walk-away. Read against the contracts your company actually signed, one of the three is routinely empty. An acceptable position nobody has ever signed is a draft wearing a tier label, and a walk-away your own deals have crossed, with no approver named against a single crossing, is not a control. So every row here carries the count of executed deals behind it, and a tier with no instances prints that instead of language.

That history is operative, not merely informative. Delaware's Uniform Commercial Code makes a sequence of conduct in previous transactions between the same two parties a course of dealing, and 1-303(d) has it supplement or qualify the terms of the agreement. The parol evidence section then lets course of dealing explain or supplement a record intended as a final expression. Its complete and exclusive statement limit sits inside paragraph (b), on consistent additional terms, not on the course of dealing in paragraph (a). Your integration clause does not switch off last year's concession.

So the register is keyed by counterparty as well as by clause type, and each position records whether the business actually operates it after signature. A term you won and never enforced is evidence of waiver under the same section, which makes an unused right a playbook problem rather than an operations one. The business team gets its own page: what to agree without asking, with the count behind each line. Applying a settled position to one inbound contract is a separate job with its own record.

Six positions, 118 signed contracts, and the three rows that did not survive the count

The Clause Position Register, the executed history behind it, and the one page the business team actually reads.

Corvidale Systems  ·  order form and master agreement

Clause Position Register

Illustrative rows for a fictional company, derived from 118 contracts executed since 2023. Enterprise 61, mid-market 57.

ClauseSegmentPreferredAcceptableWalk-away
Liability capEnterprise12 months of fees paidGreater of 12 months or 500kUncapped direct damages
Liability capMid-market12 months of fees paid18 months of fees paidAnything above 24 months
Mutual indemnityEnterpriseThird-party IP and confidentiality, inside the capSame, with a separate IP sub-capUncapped and outside the cap
Termination for convenienceEnterpriseNeither party before renewalCustomer at 12 months with a stated feeCustomer at any time, no fee
IP in feedbackBothAssigned to us, no royaltyPerpetual licence rather than assignmentRetained by the customer
Publicity and logoMid-marketNamed reference plus logoLogo on request each timeNo reference of any kind

The same six rows, against what was actually signed

ClauseDealsPref.Acc.Past walk-awayOutside all threeWhat that makes the row
Liability cap, ent.6145430Preferred is a ritual. Every deal opens by losing it
Liability cap, mid.57441300Holds. Leave it alone
Mutual indemnity61180043The middle tier has never been signed once
Termination, ent.61261142Crossed 11 times. Approver named on none of them
IP in feedback118116002Never contested. Belongs in the template, not here
Publicity, mid.5731053Four instances. Too thin to hold a position

Rows one and two are the same clause and they disagree. The preferred cap is achieved 4 times in 61 enterprise deals and 44 times in 57 mid-market ones. A single row covering both is unattainable in one segment and leaves money on the table in the other, which is why segment is a key here rather than a note.

Row three is the finding a template cannot produce. The IP sub-cap reads like a sensible middle position and has never once been signed, while 43 deals closed on language the playbook does not describe at all. Publishing that tier tells a negotiator to offer something the market has never taken.

Row four is a walk-away in name only. Eleven crossings with no approver on any of them is not a control that was overridden; it is a control nobody was applying.

Where each position came from

The enterprise liability cap, derived end to end. Every tier is a claim about these 61 deals and each one is checkable against them.

What was signedDealsMedian ACVWho asked for it
12 months of fees, no floor488kUs. All four under 100k
Greater of 12 months or a 500k floor54310kCounterparty, on their paper
24 months, no floor31.4mCounterparty, and we agreed

The preferred position is only ever achieved below 100k of annual value. That is not a negotiating position, it is a description of small deals, and publishing it as the opening ask for enterprise sends every negotiator into a conversation they lose on the first call.

The same counterparty, three times

DealSignedCap agreedTierApproved by
Westmere Health, initial2023-0624 months, no floorPast walk-awayR. Oyelaran, GC
Westmere Health, renewal 12024-1024 months, no floorPast walk-awayNot recorded
Westmere Health, renewal 22025-1124 months, no floorPast walk-awayNot recorded

Three deals, one counterparty, the same concession every time. Section 1-303(b) makes that a sequence of conduct concerning previous transactions between these two parties, and 1-303(d) has a course of dealing supplement or qualify the terms of the agreement. Renewal 3 does not open where the playbook says it opens, and a register keyed only by clause type cannot see that.

Positions won and never operated

The term we wonDeals carrying itTimes the business used it
30-day cure period before termination for breach440, across 14 qualifying breaches
Audit right on 10 days written notice380
Annual uplift capped at 5 percent, by notice513 of 51 renewals

A right nobody exercises is a playbook row, not an operations footnote. Section 1-303(f) makes a course of performance relevant to show a waiver or modification of any term inconsistent with it. Whether these three amount to that is counsel's call, and the register's job is to put them in front of counsel while the answer still costs nothing.

Business Team Guidance

The page an account executive reads. One side of paper, no clause language, and a number against every line.

If they ask forAnswerWhat you can say
A 500k floor on the liability capYes, on the call54 of our last 61 enterprise deals closed here
Mutual indemnity inside the capYes18 deals, never escalated
An IP sub-capAsk. It is written down and untested0 signed instances
A 24-month capNo. Goes to the GC3 crossings in 61, each approved by name
Termination for convenience, no feeNo. Goes to the GCUnder review. 11 unapproved crossings
Dropping the 30-day cure periodYesNever used, in 14 qualifying breaches

Deviation Log, live

DateDealClauseTier reachedApproved by
2026-02-14Aldergate MutualLiability capAcceptableWithin authority
2026-02-11Threlfall LogisticsTerminationPast walk-awayOpen, 4 days
2026-02-03Westmere HealthLiability capPast walk-awayR. Oyelaran, GC

The count is what lets a negotiator hold a line without escalating. Our policy is 12 months is a sentence the counterparty has already heard and already beaten, sometimes on a deal they signed themselves. Fifty-four of our last sixty-one closed at exactly this cap is a different conversation, and it is one an account executive can have alone.

Row two of the log is the only thing on this page that needs a human today. The rest closed inside somebody's authority and recorded itself.

What's in the pack

01

Clause Position Register

Preferred, acceptable and walk-away per clause per segment, each tier carrying the count of executed deals that landed on it and the ones that landed outside all three.

02

Executed History

One row per contract per clause: what was actually signed, which tier it reached, the counterparty, the deal size and who approved anything past the walk-away.

03

Business Team Guidance

One side of paper for a non-lawyer. What to agree on the call, what goes upstairs, and the number behind each line that makes it sayable in a negotiation.

04

The Playbook

The clause-by-clause document with the rationale for each tier, written so the position and the evidence for it arrive in the same paragraph.

05

Deviation Log

Live departures as they happen, with the tier reached and the approver named. A crossing with nobody against it stays open and reads as open.

06

Approval Matrix

Who signs off what, by tier and by threshold, with roles rather than people, so a departure routes itself instead of finding the nearest lawyer.

07

Reading Your Own History

The method doc: how a tier gets derived, what counts as an instance, and why a position with no instances stays marked untested rather than being published.

08

What The Playbook Does Not Cover

The clause types with too few executed instances to hold a position, named individually. Applying these positions to one inbound contract is the next job.

How to use it

  1. 1

    Open in River, or take it blank

    Open the pack in River and hand it your executed contracts, or download the Word documents and CSV sheets from the template library and fill them in yourself.

  2. 2

    Send the contracts you signed

    Executed agreements, order forms and amendments, in whatever shape they sit in. The final signed version is the evidence, so early drafts are noise here.

  3. 3

    Every tier gets counted

    Each clause type resolves into what was actually agreed, how often, in which segment and with which counterparty. Tiers with no instances are marked untested rather than filled in.

  4. 4

    A named person publishes it

    The counts are evidence; the position is a decision. Each row takes an owner and a date before it reaches the business team's page.

Frequently asked questions

Is this template free?

Yes. The zip is Word documents and CSV sheets, no account and no card. Edit with AI is the other half: the agent reads your executed contracts, counts what each clause actually settled at, and marks every tier your own history cannot support. Other packs sit in the template library.

What format are the downloaded files?

Word (.docx) for the five documents and CSV (.csv) for the four sheets, zipped together. Excel, Numbers and Google Sheets open the registers straight off the download, and the playbook and guidance pages open in Word or Pages. Nothing to convert.

How many executed contracts does this need?

Enough that a count means something, and the register tells you which rows are not there yet. A clause with four instances across fifty-seven deals is marked too thin to hold a position rather than given one, because a tier derived from four deals is a guess with a number stapled to it.

Why is the counterparty a column and not a note?

Because repeated conduct with the same party is legally operative. Under Delaware's course of dealing section, a sequence of conduct concerning previous transactions between those parties may supplement or qualify the terms of their agreement. A register keyed only by clause type cannot see the third identical concession to the same customer.

Our contracts all have an integration clause. Does that not settle it?

It settles less than it reads like it does. The parol evidence section puts its complete and exclusive statement limit on consistent additional terms in paragraph (b), while paragraph (a) lets course of performance, course of dealing and usage of trade explain or supplement the record regardless. Your history travels with the merger clause in place.

How is this different from reviewing a contract against a playbook?

This builds the positions; that applies them. Once a position set exists, reading one inbound contract against it and recording who called each issue is a separate job, and an NDA needs one more test: which side each clause helps here. Once a deal is signed, the obligations it created belong in a standing register, which is also where the never-operated rights surface.

What does 'Edit with AI' actually do?

It signs you up, installs this exact pack as a private workspace, and puts the agent in front of an empty Clause Position Register. Then you send your executed contracts and whatever guidance already exists, including a playbook you think is out of date, which is the most useful thing you can send.

Build the playbook out of what you actually signed

Take the Word documents and CSV sheets blank, or open this exact pack in River and hand it your executed contracts.

Edit with AI