Choose work with a reason
Choose the next change from the evidence.
A short task can still be the wrong task. Before choosing what comes next, connect the proposed change to a buyer question, inspect its source and record why the work deserves the effort. Our account queue keeps that decision beside the evidence.
By the Broadcastwell team. Updated 7 November 2026.
Agree what value means before assigning a number
The queue includes fixes from delivered work and cited pages your team has begun to work on. Each item retains its owner, buyer questions, source page, receipt and dated state. Open that evidence before making an estimate. An empty queue means there is no retained work to rank.
Use the same business objective and rating scale across the work you compare. A task's goal should explain how it supports that objective. A useful goal names the buyer decision or information gap that the proposed page change addresses.
- Expected value: choose 1 for a small contribution to the agreed objective, through 5 for a major contribution. Explain the basis in the supporting note.
- Estimated effort: include research, drafting, review and publication time. Enter 0.25 to 2,000 hours. The estimate describes total work, not a promised delivery date.
- Confidence: choose 1 when supporting evidence is weak, through 5 when it is strong. Explain what remains uncertain. This rating is not a statistical confidence level or a probability.
- Supporting evidence and assumptions: refer to the retained question, source and receipt. Name any dependency, missing source detail or outside decision that could change the estimate.
People with write access can save an assessment. Readers can inspect it and its history. A team member's access to one brand gives no access to another brand's queue.
Read the calculation before following the rank
The estimate score is expected value multiplied by confidence, divided by five times estimated hours. Higher scores appear first among assessed work. The score is a sorting aid made from your estimates. It is not measured lift, revenue, a forecast or proof that a change will affect an answer.
For example, value 5, confidence 4 and effort 3 hours produce 20 divided by 15, or 1.333333 when displayed. Two tasks with the same score keep their existing priority order, then their latest recorded date and stable identifier. This makes the order repeatable.
Unassessed work follows the assessed queue and has no invented score. Published, measured and declined items remain below the work queue with their evidence and earlier estimates. Every item stays in the record. The state filters and column controls help you inspect a subset; the complete exports retain all items.
If assessment storage is unavailable, the page says so and prevents a save. Existing work remains visible. A missing estimate does not mean the task has no value.
Follow one complete fictional decision
Kalvenor is fictional (Sample data, fictional company)
The sample queue contains seven retained changes. Its companies, answers, owners and work history are fictional. No work in this example was delivered to a client.
- Open the procurement checklist change. Follow its receipt and source context for questions 18 and 25. The proposed work concerns contract terms and data export detail.
- Expand Assessment and revisions. Its fictional reviewer assigned value 5, confidence 4 and 3 hours of effort. The rationale connects those inputs to the missing detail. Its displayed estimate score is 1.333333.
- Compare the field-note task for question 30. Value 3, confidence 3 and 4 hours produce 0.45. The finance buyer task has no assessment, so it remains visible without a score.
- Inspect the earlier published and measured records below the active work. Their dates remain the dates of the retained observations and fictional work. Saving a new assessment does not change those dates.
- Export the CSV or open the Markdown. Find the same question identifiers, receipt, state history and complete assessment revisions.
The sample is read only. In your account, choose Assess work beside an active item, fill in the inputs and save. Open the record again to inspect the saved revision and new position in the queue.
Keep earlier judgments when an estimate changes
A revised estimate adds a new record. It retains the earlier goal, ratings, effort, explanation and recorded time. If another person saves first, your older form cannot replace that decision silently. Reload the queue, read the newer revision and decide whether another change is warranted.
An assessment does not approve the proposed edit, change its owner, publish a page, send a message or start a measurement. Use the existing review and publication record for those decisions. The later observation remains separate from the estimate that helped choose the work.
Once an item has a publication, measurement or declined record, its assessment history remains readable. New work should have its own proposed change and evidence rather than rewriting the judgment attached to completed work.
Carry the complete queue into the next conversation
The ledger's CSV retains work rank, queue group, value, effort, confidence, score and every assessment revision beside the existing fields. The Jira CSV and Markdown include the same history. These are files for your review or import; creating them sends nothing to another service.
Your complete account archive includes all retained assessment revisions. The existing read-only fixes responses also include the implementation queue. In a connected assistant, ask for get_fixes for an existing delivered run; the returned queue is limited to that run. The account fixes API includes the account queue, while its run-specific response includes only that run's entries. Existing access rules apply.
Review the underlying receipt before presenting a ranked task as a recommendation. The rank records a choice about work. A claim about results needs the observations and comparison that support it.
Inspect the fictional queue, open your account's Fixes view or read the measurement export guide.
Once you choose the next change, assign a person and record the handoff. The assignment keeps the estimate and evidence intact.