Method note: this brief proposes a source-led review approach. It does not claim that private client or firm data has been collected.
Question and scope
This protocol asks how much of the prebill review effort is spent locating source records rather than evaluating the billing substance. Locating includes searching for the matter record, the client guideline, the time entry, the approval, or the supporting document. Substantive review includes reading the evidence and deciding whether the entry is complete and appropriate. The distinction matters because search time is often hidden inside a general review estimate.
Define the matters, reviewers, period, and review stage included in the measurement. A prebill review can involve an assistant, a billing manager, and a partner, and each role may spend time differently. Record the role and the stage for every observation. The question is descriptive, and it should not be used to compare individual reviewers without accounting for the records they were given.
Unit of analysis and observation
Use one review episode as the unit, from the moment the reviewer opens the prebill to the moment the review is recorded as complete. If the review is interrupted and resumed, keep the elapsed work time separate from the wall-clock time. A short interruption should not make a routine review look inefficient. The observation should capture the start state, the end state, and the reason for any pause.
Record the matter, the reviewer role, the number of lines, and the systems used. These variables help explain why one review took longer than another. A large matter with a new client may legitimately need more search time than a repeat matter with a stable record. The protocol should describe the work, not rank the people doing it.
Defining search time
Search time begins when the reviewer starts looking for a record and ends when the record is found or the reviewer decides it is unavailable. It includes opening systems, checking messages, asking a colleague, and comparing versions. It does not include reading the record once found or writing the review note. The boundary should be written down and applied consistently so two observers classify the same activity the same way.
Some activities blend search and review, such as reading a guideline to find the applicable rule. Classify the activity by its primary purpose and record the ambiguity rather than forcing a false precision. If a reviewer spends time deciding whether a document is the right version, that is search until the version is confirmed. A short decision rule at the top of the observation form reduces disagreement.
Evidence and data collection
Collect the observation with a simple form that records the start time, each search segment, the source sought, and the outcome. Ask reviewers to note the system and the reason for the search in a few words. Do not require a stopwatch for every task if the interruption would distort the work; a short retrospective estimate can supplement direct observation. Record which method was used for each episode.
Keep the observation data separate from the review record so the measurement does not change the work being measured. If the team is uncomfortable with direct timing, start with a sample of self-reported episodes and state the method limitation. A consistent method matters more than a sophisticated one. The goal is a useful estimate of where effort goes, not a billing audit of the reviewer.
Sampling and stratification
Collect episodes across matter types, client guidelines, reviewer roles, and record ages. Search time is likely to be higher for new matters, incomplete records, and clients with long guidelines. A sample drawn only from repeat matters will understate the problem. If the population is large, stratify and report the counts for each stratum rather than a single average.
Include episodes where the source was never found, because an unavailable record can consume more time than a successful search. Record the reason it was unavailable and the escalation that followed. Excluding these episodes would make the search problem look smaller than it is. The report should show both the successful and unsuccessful searches. An unavailable record that was never created is a different finding from one that exists but cannot be located.
Analysis and reporting
Report the median and range of search time per episode, the share of total review time, and the sources that took the longest to locate. A median is more robust than an average when one difficult matter dominates the sample. Show the number of episodes and the method used to collect them. Avoid presenting a precise percentage when the sample is small or the method is retrospective.
Group the findings by source type so the owner can see whether the problem is the matter record, the guideline, the time entry, or the approval trail. For each group, describe the current location and the access barrier. A finding that says reviewers spend time searching is less useful than one that says the prior approval is stored in a personal inbox. The report should point to a specific fix.
Limits and sources of error
Direct observation can change behavior, and self-report can undercount short searches. Reviewers may disagree about where search ends and review begins, especially when a document is hard to interpret. The record itself may be incomplete, and a missing document can be recorded as a search failure even when it was never created. State these limits with the result.
The cited public sources frame recordkeeping, privacy, and internal control practice; they do not provide a time estimate for the firm. The protocol measures the observed episodes, not the value of the review or the quality of the decision. Do not use the result to judge individual performance without adjusting for the records and systems each reviewer received. The purpose is to improve access to evidence.
What counts as substantive review
Substantive review begins when the reviewer has the record in front of them and starts evaluating the entry against the standard. Reading a guideline to confirm the applicable rule can be part of that evaluation, but searching for the guideline is not. Write examples for the team so the boundary is not left to individual interpretation. The distinction should be stable enough to compare episodes across reviewers.
A fast review is not automatically a good review, and a slow review is not automatically a thorough one. The time measure should be paired with a small quality sample so speed does not become the only goal. If search time falls while defect rates rise, the access change may have hidden the evidence rather than improved it. Report both dimensions when the owner is deciding whether to keep a change. The quality sample should use the same checklist before and after the pilot.
Testing an access improvement
Pilot one access change with a single client or matter type, such as a linked source panel or a standard evidence folder. Measure search time and review quality before and after the change on comparable episodes. Ask reviewers whether the new location actually reduced the number of places they had to check. A change that adds another system without removing a search step may not help.
Keep the pilot small enough to reverse if the evidence quality drops. Compare a sample of reviewed entries against the same checklist used before the change. If the new access path makes the right document easier to find, both search time and consistency should improve. Record the result and the date before expanding the change.
Review cadence
Repeat the measurement after a record-location change, a system migration, or a new guideline template. Compare the same roles and source types so the movement is meaningful. Set a review date and name the operations owner who will act on the result. A measurement that is not repeated will not show whether the fix worked.
Choose one access improvement and test it on the next sample. The improvement might be a linked source panel, a standard folder structure, or a matter-opening field that stores the guideline reference. Keep the observation forms and the analysis notes together so a later reviewer can repeat the method. The measure earns its place when it reduces search time without reducing review quality.
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


