Home / Research / Billing guidelines

How often do client billing guideline exceptions repeat across matters?

A measurement protocol for counting repeated client guideline exceptions across matters, with definitions, sampling rules, evidence standards, and reporting limits.

Legal Billing Assistant Editorial Team · Published

Editorial illustration representing How often do client billing guideline exceptions repeat across matters?
Table of contents

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 often the same client billing guideline exception appears across different matters in a defined billing period. The purpose is to separate a one-off correction from a rule that keeps producing the same problem. The unit is an exception event, not a time entry and not a matter, because one matter can generate several distinct exceptions. The result is a description of the observed population, not an audit opinion or a claim about the client's intent.

The scope should name the client or client group, the matters included, the billing period, and the guideline version in force. A protocol that mixes clients with different guidelines will produce a repeat rate that cannot be interpreted. Record the source document and the date it became effective before collecting any exception. The question is narrow enough to answer with records, and it should stay narrow when the queue is busy.

Population and unit of analysis

Define the eligible population as every prebill line or invoice reviewed under the named guideline during the window. State whether lines, invoices, or matters are the denominator and keep that choice consistent throughout the analysis. A line-level rate answers a different question from a matter-level rate, and both can be useful when labeled correctly. Never switch the denominator midway because the first result looked unhelpful.

Count an exception event when a reviewer identifies a guideline departure and records the rule, the evidence, and the disposition. If the same line is reopened and reclassified, keep the original event and record the later disposition as a linked update. A reopened item is not a new exception unless it concerns a different rule or a different line. This rule prevents the same problem from being counted twice and inflating the repeat rate.

Defining a guideline exception

Write the exception definition as a test that two reviewers can apply to the same record. For example, a block-billing exception exists when a narrative combines two or more distinct tasks in a way the client guideline prohibits. The definition should name the rule, the field inspected, and the evidence that confirms the departure. A definition that depends on the reviewer's general impression will not support a repeat count.

Distinguish a guideline exception from a data error and from a billing judgment. A missing matter number is a data error, while a rate that exceeds the agreed schedule may be a guideline exception with a separate approval path. A decision to reduce a charge for relationship reasons is a judgment, not a guideline departure. Keep these categories separate so the repeat rate measures the rule it claims to measure.

Evidence standard

Every counted exception needs the source record, the guideline reference, the observation date, and the reviewer who made the call. A screenshot taken after the prebill changed cannot prove the original state, so record the version or timestamp observed. Store the evidence under a pseudonymous case key with the crosswalk in a restricted system. The evidence standard should let a second reviewer reconstruct the decision without contacting the first.

Mark missing evidence separately from a confirmed absence. If the guideline cannot be located, the item is unresolved rather than compliant. If the record shows no exception after a complete check, that is a confirmed negative and belongs in the denominator. Collapsing these states hides the difference between clean work and unexamined work. The report should show the unresolved count beside the exception count.

Sampling and collection

Collect consecutive reviewed items where possible, then stratify by matter type, timekeeper, and review owner. A sample drawn only from the easiest matters will understate the repeat rate and mislead the improvement effort. If the population is too large for a full count, state the sample size and the selection rule. Include adverse cases such as a missing guideline, a conflicting instruction, and a late correction.

Record the collection window, the eligible count, the number reviewed, and every exclusion with its reason. A reader should be able to reproduce the sample from the protocol and the source list. If the guideline changed during the window, split the population at the effective date instead of blending versions. A clean collection log is more valuable than a large sample with an unclear origin.

Coding repeat exceptions

Create a short code list before coding begins and define each code with an example and a non-example. Codes might cover narrative structure, travel time, administrative work, rate application, and documentation. Apply one primary code per exception and allow a secondary code only when the rule genuinely overlaps. Too many codes make the pattern difficult to see, while too few hide the specific rule that needs attention.

Have a second reviewer code a small overlap sample and compare the results. Disagreement is evidence that a definition or an example is weak, not necessarily that a reviewer is careless. Resolve the disagreement by clarifying the codebook before coding the rest of the population. Record the change and its effective date so the earlier and later results can be interpreted honestly.

Analysis and reporting

Report the eligible count, the reviewed count, the exception count, and the repeat count by code and by matter. Show the unresolved and missing-evidence counts beside the main result rather than in a footnote. If the sample is small, report the raw counts and avoid a precise-looking percentage. A simple table with clear denominators is more useful than a polished chart that hides the population.

Separate observation from interpretation in the report. The observation is that a code appeared in a certain number of matters; the interpretation is that the guideline summary or the intake check may need repair. Give the accountable owner the decision the report supports, such as revising a template or adding a review checkpoint. Do not imply that the repeat rate measures a person's performance.

Limits and sources of error

Small samples are sensitive to one matter, and a repeat rate can move sharply when a single large client enters or leaves the window. Reviewer disagreement, changing guidelines, and incomplete records can also distort the result. The report should state these limits next to the finding rather than leaving them to the reader. A limitation that is not written down tends to disappear when the result is summarized.

The public sources cited in this protocol frame recordkeeping, privacy, security, and internal control practice; they do not provide a result for the firm's sample. The method cannot infer missing work from a clean summary or prove that an unrecorded exception did not occur. Use the result to improve the operating system, not to judge professional worth or to replace client-specific advice. A cautious reading is part of a credible measurement practice.

Retention and privacy

Keep the codebook, the sample list, and the evidence references for the period named in the firm's retention policy. Store the evidence under a pseudonymous case key and keep the crosswalk in a restricted system. Do not place client-confidential narrative in a shared notes file that is broader than the review team. A repeat study is useful only when the records behind it remain readable and appropriately protected.

When the guideline or the sample changes, archive the prior version rather than overwriting it. A later reviewer may need to know which rule applied to an earlier exception. Record the date the version was retired and the person who approved the change. Retention is part of the method because a result without its source cannot be reproduced.

Applying the result to the guideline summary

Use the most frequent repeat codes to decide which parts of the guideline summary need a clearer example or a new check. A code that appears across several matters is a stronger candidate for a template change than a single difficult entry. Ask the review owner to confirm the interpretation before changing the summary. The goal is a smaller repeat rate in the next window, not a longer list of rules.

Add one prebill checkpoint for the leading code and assign an owner to test it. Run the same protocol after the change and compare the counts by code and matter. If the repeat rate does not move, the checkpoint may be addressing the symptom rather than the cause. Record what was tried so the next reviewer does not repeat the same experiment.

Review cadence

Set a review date for the result and name the person who will read it. Re-run the protocol after a guideline change, a template change, or a shift in the client mix, because the earlier rate may no longer describe the queue. Compare the same definitions and report what moved, what did not, and what remains unknown. A report without a review owner becomes an archive rather than a control.

Choose one next action from the review: fix the guideline summary, simplify a field, add a prebill check, or stop the work until the owner decides. Assign the action and a date, then test whether the repeat rate changes in the next window. Keep the protocol, codebook, sample file, and open questions together so a later reviewer can repeat the method. The value of the exercise is the improvement it prompts, not the number alone.

Authoritative sources

These public references inform the operating questions in this article. They do not replace legal, ethics, privacy, financial, or client-specific advice.

  1. American Bar Association Model Rules of Professional Conduct
  2. NIST Privacy Framework
  3. NIST Cybersecurity Framework
  4. U.S. Government Accountability Office Green Book
  5. Internal Revenue Service recordkeeping guidance
  6. Legal Electronic Data Exchange Standard

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