Start with the output the role owns
A billing support role produces review-ready records, not activity. Start the scorecard with the outputs the role is accountable for, such as a complete prebill packet, a reconciled cost log, or a rejection queue with named owners. Describe each output in a way a reviewer can check without watching the person work. If the output cannot be defined, the role boundary is still unclear.
Separate the outputs the role prepares from the decisions an attorney or billing manager owns. The scorecard should measure the quality of the preparation and the timeliness of the escalation, not the outcome of a judgment call. A good scorecard makes that distinction visible. A scorecard that rewards a favorable write-off decision will push the assistant toward work the role should not do.
Measure completeness before volume
Count the records that contain every required field before counting how many records were touched. Volume without completeness creates rework downstream and hides the real cost of the role. Use a short field list tied to the workflow, such as source, owner, date, amount, and next action. Report the complete share beside the total so a high volume cannot mask a low quality rate.
Define what partial and missing mean before the first scoring period. A record with a source but no owner is partial, while a record with no source at all is missing. Keep the definitions in the scorecard guide so two reviewers apply them the same way. The measure should point to a specific repair, not just a percentage.
Score the evidence, not the activity
A scorecard can count logins, messages, or entries, but those numbers say little about whether the work can be reviewed. Replace activity counts with evidence checks, such as whether the packet includes the client authorization or whether the rejection note explains the correction. Ask the reviewer to sample the evidence rather than rely on a self-report. The sample can be small if the definition is clear.
Use a simple scale, such as complete, partial, or missing, and keep it consistent across periods. Avoid a weighted score that hides a serious gap behind several minor successes. If a record lacks the one document that proves authority, the score should reflect that failure plainly. The role holder needs to know which gap matters most.
Add a rework and exception measure
Measure how often a prepared record comes back for correction and what caused the return. Group the causes into missing evidence, wrong matter, unclear instruction, system issue, and judgment question. A high rework rate may point to the handoff, the template, or the training rather than to the person. Report the cause distribution with the rate so the team can choose a fix.
Track exceptions that were escalated correctly as a positive sign, not a failure. A role that never escalates may be guessing, while a role that escalates everything may lack decision rights. Review a sample of escalations to see whether the evidence and the question were clear. The scorecard should reward an early, well-documented escalation.
Keep client and matter privacy in the scorecard
Score at the level of the queue, the task type, or the role, not the client name. Remove amounts, matter descriptions, and personal details that are not needed for the review. If an example is used in coaching, redact it and use a role label. The scorecard is a management tool, not a second client record.
Limit access to the underlying sample and keep it in a controlled folder. Aggregate before sharing results beyond the manager and the role holder. Check that a small group cannot be re-identified from a chart or a table. Privacy rules apply to internal performance data just as they apply to client files.
Calibrate with a second reviewer
Have a second reviewer score the same sample before the scorecard is used for a formal review. Compare the results at the record level and discuss every disagreement. If the reviewers interpret a field differently, revise the definition rather than averaging the scores. Calibration is especially important in the first two periods.
Record the agreement rate and the changes made to the guide. Repeat the calibration after a template or system change, because the old definitions may no longer fit. A scorecard that drifts between reviewers is not a fair measure. The calibration note should be part of the scorecard's own evidence.
Review the scorecard with the role holder
Give the role holder the sample, the definitions, and the result before the review meeting. Ask which gaps are caused by missing access, unclear instructions, or an unrealistic deadline. Listen for the workarounds that the scorecard does not capture. The conversation should produce one process change and one support change.
Separate coaching from evaluation and be explicit about which is happening. Use the scorecard to improve the workflow, not to rank people against each other. If a measure consistently punishes a role for a system problem, remove or redesign it. A scorecard that the role holder cannot influence will be ignored.
Retire measures that reward the wrong work
Review the scorecard each quarter and remove any measure that encourages speed over evidence. Watch for gaming, such as closing records without the required source or escalating routine work to avoid a quality check. Compare the scorecard with the actual downstream rework and client questions. A measure that looks good while rework rises is measuring the wrong thing.
Keep a short history of the measures and the reasons they changed. Tell the role holder what changed and why so the revision does not feel arbitrary. Add new measures only when the team can define them and review them consistently. The best scorecard is small, stable, and tied to the output the role actually owns.
Common questions answered
Who owns the scorecard? The billing manager owns the scorecard and the review conversation, while the role holder helps define the evidence and the constraints that affect the result. The billing manager owns the review and the role holder helps define the evidence. Both should agree on the constraints before the first scoring period.
What should a scorecard measure first? Start with the completeness of the output and the rate of rework, then add exception handling and timeliness once the definitions are stable. Begin with completeness and rework, then add exception handling once the definitions are stable. A small, stable set of measures is easier to calibrate.
How is fairness protected? A second reviewer calibrates the sample, definitions are written down, and aggregate results are reviewed with the role holder before any formal evaluation. A second reviewer checks the sample and the definitions are written down. Aggregate results are discussed with the role holder before any formal evaluation.
Put the guidance to work
Choose one recurring billing workflow, identify its authoritative source records, and name the owner who reviews exceptions. We can help turn that boundary into a focused support role.
Discuss your workflow


