Cross-Team Dependency Negotiation Template
Two documents and three sheets that check a provider team's stated confidence against what that same confidence level has actually meant before.
Free download · No account needed
Dependency Register, one row
What was asked, what was promised, and what that promise has meant before
Filled in once a commitment is negotiated and the provider's own Slip History by Team sheet has enough past commitments to read.
Stated Confidence
What the provider team says right now, as a number or a mapped equivalent: "90 percent confident," "should be fine," "no promises."
Historical Rate At That Confidence
Pulled from Slip History by Team: of everything this team has said at this same confidence level before, how much actually landed on time.
Gap
Stated minus historical. Under 10 points either way reads as calibrated. Twenty or more reads as an overconfidence flag, regardless of how firmly the date was stated.
Check-in Cadence
Weekly once the gap reaches 15 points in the overconfident direction. Biweekly otherwise, including when a team understates itself.
Every cross-team dependency tracker ranking for this query records the same three things on each row: a status of on track, at risk or blocked, a provider-stated confidence or risk score, and an owner on each side. None of them check that confidence against anything. The UK government's own reference-class-forecasting research draws exactly the line these trackers miss. An inside view, built from the specifics of one commitment, is where the optimism comes from. Only an outside view, built from a comparable track record, actually corrects it.
This pack builds that outside view for every provider team it tracks. Past commitments get bucketed by the confidence a team stated at the time, the same check forecasters run on their own published probabilities. A stated 70 percent should hold about 70 percent of the time, measured across many claims made at that level. At Thornwood Systems, a fictional healthcare scheduling software company, Identity's stated 90 percent confidence has held on 3 of its last 12 commitments at that level, 25.0 percent, its worst-calibrated band of the three it uses.
Across all three of Thornwood's provider teams, 120 recorded commitments held on time 69.2 percent of the time, a figure no single dependency row would surface. Slip history by team reads as uncomfortable, and it is exactly what tells a requesting team which commitments to plan around rather than believe. Built for the lead negotiating a date out of a team they do not manage. The technical program plan tracks the same handoffs at the programme level, and the capacity forecast checks a provider's own backlog before it commits to anything.
What's in the pack
Dependency Agreement
The five fields a real commitment needs, including the provider's historical rate at its own stated confidence, attached at the moment the commitment is recorded.
Escalation Note
Fires on a missed check-in, a downward confidence revision, or a historical rate already under 50 percent, and gives three concrete options instead of a request to watch it.
Slip History by Team
Every provider team's past commitments bucketed by the confidence stated at the time, with the actual on-time rate and the gap for each band.
Dependency Register
Every open ask, the provider's fresh confidence next to their historical rate at that level, and the resulting status: calibrated, watch, flag or conservative.
Check-in Schedule
Cadence set from the size of the calibration gap, weekly for the least-calibrated confidence claims, with whether each check-in actually happened on schedule.
How it works
- 1
Open in River, or take it blank
Claim the pack in River and hand it your commitment history, however informal, or take the blank documents and sheets and fill them in yourself.
- 2
Send the commitment history
Past cross-team asks and whatever confidence or status language each provider team used, plus what actually happened by the committed date.
- 3
Get the calibrated register
River bucket-sorts each team's history by stated confidence, computes the real on-time rate per band, and checks every open ask against it.
- 4
Negotiate with the record in hand
Set check-in cadence from the gap, and escalate with the historical rate stated plainly rather than a guess about the relationship.
Frequently asked questions
Is this template free?
Yes, no account or card needed to download it. Edit with AI is the optional half, where River reads your commitment history and builds the calibrated register itself. Every other pack sits in the template library.
What format are the downloaded files?
Two documents as Word files and three sheets as CSVs, zipped together. The sheets ship with Thornwood Systems' illustrative commitment history already in place, so the calibration gap is visible before you replace the rows with your own teams' data.
How is this different from a status of on track, at risk or blocked?
Those three words describe how this one commitment feels right now, from the provider's side. This pack adds what the same provider's confidence has actually meant across its past commitments, which is the only way to tell a well-calibrated 90 percent from one that has held a quarter of the time.
What if we don't have enough commitment history yet?
The sheet says so directly, as insufficient history, rather than reporting a rate off two or three data points. Until five or more commitments exist at a given confidence band for a team, that band's row stays marked unproven instead of quietly averaged into a number nobody should trust yet.
Does this replace our technical program plan or exec update?
No. The technical program plan runs the whole programme's dependency network and shock-tests it; this pack is where one specific ask gets negotiated and checked. The engineering status update reports the outcome upward once it is set.
Our commitment data can't leave our network. Can we still use this?
Yes. A private AI workspace builds the register, the check-in schedule and the slip history inside your own tenancy. No team name, date or confidence level crosses a boundary your security team has not already reviewed and approved for internal delivery data like this.
Check a promise against what it has actually meant before
Send the commitment history you have. River builds the calibrated register, sets the check-in cadence, and tells you which teams' confidence to plan around rather than believe.
Edit with AI