Cross-Team Handoff Agreement Template
Four documents and five sheets that turn your rejection history into a definition of done, so a bounced handoff produces evidence instead of an argument.
Free download · No account needed
Handoff Checklist
Pennard Systems, Sales to Implementation
| Derived from | 214 handoffs, 68 rejection events, two quarters |
| Candidates priced | 9, tested against the check each one would impose |
| Adopted | 4, twenty-one minutes total, owned by the account executive |
What must be true before the deal is marked closed-won
| Code | What has to be true | Evidence | Min |
|---|---|---|---|
| SI-1 | Signed order form lists the modules sold | Form, module lines visible | 3 |
| SI-2 | Named client sponsor with sign-off authority | Name, title, email, confirmation | 4 |
| SI-3 | Pay frequency and pay group count recorded | Both values in the deal record | 6 |
| SI-4 | Go-live date confirmed by implementation first | Dated reply naming the date | 8 |
Five more causes fired in the same log. Three became required fields on the order form, one became implementation's own task, and one is accepted at half a day a quarter: none of the five may bounce a handoff.
Every handoff template on the market hands you the same blank field: write your acceptance criteria, filled in from a meeting where both teams remember the last argument differently. A definition of done built that way survives exactly as long as the next disagreement. Once a process map or an ownership matrix has settled where one team's part ends, this pack settles what 'ended completely' means at that exact seam. It's derived from what the receiving team has actually sent back, not from what either side remembers.
Every rejection cause in the history becomes a candidate, adopted only when the rework it prevents beats the cost of checking every handoff for it, the same cost-benefit test Executive Order 12866 has required of federal rule-making since 1993. The Rejection Log is the input, and the Definition of Done Economics sheet runs the test on every candidate. The Handoff Checklist holds what survives, and Cycle Time by Handoff prices the result in days, with the checking time added back in rather than hidden.
Pennard Systems, a payroll and HR vendor, ran the derivation across two crossings and two quarters: 390 handoffs, 99 rejection events, 13 candidates, 7 adopted. Twenty-one minutes of checking on the Sales to Implementation crossing removed 262 of the 370 calendar days bounces had added, and mean time to kickoff fell from 4.83 days to 3.65. The agreement also borrows a verdict from conditional acceptance in federal procurement: the receiving team starts the work anyway, and the sender still owes one named item by a set date.
What's in the pack
Rejection Log
Every bounced handoff for two quarters and two crossings, with the cause code, the calendar days it added, and who bounced it, kept as one row per event rather than per handoff.
Definition of Done Economics
Prices every candidate cause by the rework it avoids against the check it would cost on every handoff, then rules: in the definition of done, a form field instead, the receiving team's own work, or accepted outright.
Handoff Checklist
Only the items that survived the test, plus the ones that failed it, flagged so nobody mistakes a required form field for a reason to bounce a handoff.
Cycle Time by Handoff
Mean days from tender to acceptance, clean against bounced, with the checking time added back in rather than left out of the after picture.
Dispute Register
What a receiving team said when it bounced work back, whether the agreement covers it, the ruling, and whether the dispute found a defect in the work, the agreement, or nothing at all.
Handoff Agreement
The signed document for both crossings: the definition of done, the three responses to a tendered handoff, and what a bounce has to name. An escalation matrix covers what breaks; this covers what's merely incomplete.
Dispute Note
The ten-minute procedure for a contested bounce, naming who rules on it and the difference between a defect in the work and a defect in the agreement's wording.
How the Definition Is Derived, and Its Limits
The four-figure test behind every adopted item, plus what it cannot do: find a cause nobody logged, run on a crossing that happens twice a year, or settle a genuine scope dispute.
How to use it
- 1
Open in River, or download it
Install the pack and hand River your rejection history, or download all four documents and five sheets blank and run the derivation yourself.
- 2
Send the rejection history
A queue export with status changes, a shared inbox, the channel where work gets bounced, or just the handoff count for the period and the causes the receiving team is sick of chasing.
- 3
Price each candidate cause
For every rejection cause, name roughly how many minutes a check would cost on every handoff and how many minutes the rework actually costs when it happens, then let the test decide.
- 4
Sign the checklist, log disputes
The surviving items become the Handoff Checklist both sides work from, and every contested bounce afterward goes in the Dispute Register instead of back into an argument.
Frequently asked questions
Is this free, and what do I get?
Free, with no signup on the download: four documents and five spreadsheets. The AI half is optional. Send River your rejection history and it derives the definition of done and prices every candidate cause for you.
Why not just write the acceptance criteria in a meeting?
Because a list written from memory is always too long. A cause that fires on 6 of 214 handoffs still imposes its check on all 214 of them, and that ratio is invisible to anyone sitting in the meeting where the list gets written.
What happens to a rejection cause that fails the cost test?
It does not disappear. Depending on what it actually needs, it becomes a required field on the sending team's form, moves to the receiving team as their own task, or is accepted outright with the arithmetic written down next to it.
How is a disputed bounce different from an ordinary one?
An ordinary bounce cites a code from the checklist and gets fixed. A disputed one is ruled on by whoever owns that crossing, in about ten minutes, and the ruling records whether the defect was in the work, the agreement's wording, or neither.
Does this work for a handoff that only happens a few times a year?
Below a few dozen handoffs a year, the pack has you skip the arithmetic and write the definition of done from what you remember failing. Six rejections against six handoffs is an anecdote, not a rate the test can run on.
What does Edit with AI actually do?
Creates a free account, installs this exact pack as a private space, and primes the agent to read whatever rejection history you send. Nothing gets written until you send something, and the blank pack is always there to download instead.
Stop settling the same bounce twice
Take the four documents and five sheets blank, or install this pack in River and send your rejection history.
Edit with AI