You can use Billables AI Intelligence feature to write your firm Billing Guidelines. Follow these simple steps and you'll be up and running.

You are helping our firm write its organization-level billing guidelines for

Billables AI.

Below are example time entries from our approved, billed work product. They

represent how we want our time entries to read.

Work through the five parts below in order.

───────────────────────────────────

PART 1 — ASSESS THE SAMPLE FIRST

Before inferring anything, report:

- How many distinct timekeepers the entries come from

- How many distinct clients, matters, and practice areas

- What date range the entries span

- How many entries in total

Then state plainly how much weight the sample can carry. In particular:

if the entries come from one or two timekeepers, say so, and treat any

stylistic pattern (verb choice, sentence shape, level of detail) as that

person's habit rather than firm policy. Present those as candidates for

firm review, not as rules.

──────────────────────────────────────────────────────────────────────

PART 2 — WHAT THE EXAMPLES SHOW

Infer the conventions these entries follow. Cover, where the evidence

supports it:

- Sentence structure and opening

- Tense and voice

- Punctuation and capitalization

- Level of detail: task-level vs. play-by-play; whether background facts appear

- How people are named: full names, initials, role descriptions, opposing parties

- Preferred and avoided verbs

- Whether several activities appear in one entry or each gets its own

- How identifiers appear (Bates ranges, policy numbers, property addresses)

- Any convention specific to one client or matter that differs from the rest

───────────────────────────────────

PART 3 — WHAT THE EXAMPLES DO NOT SHOW

The absence of a category of work is often firm policy, not an oversight.

For each item below, say whether it appears in the sample. If it does not,

propose the guideline that would match that absence:

- Internal email between firm personnel

- Internal meetings, status calls, staffing discussions

- Client lunches, dinners, travel, events

- Administrative time (timekeeping, conflicts checks, file opening, billing)

- Email read but not replied to

- Legal research platforms as a standalone entry

- Document review at a per-document level

- Short calls and voicemails

- Personal calendar events and blocks

Label each one:

CONFIRMED — the sample directly demonstrates the convention

LIKELY — the pattern is consistent across the sample

UNCERTAIN — the sample is silent and this could go either way

Only CONFIRMED and LIKELY items may become guidelines. Every UNCERTAIN item

goes in the review section as a plain question.

───────────────────────────────────

PART 4 — CHECK WHAT BILLING GUIDELINES CAN ACTUALLY CONTROL

This is the most important step. Organization-level billing guidelines are a

free-text box that shapes how Billables AI writes and filters time entries.

They govern only:

- What a narrative says and how it is worded

- What detail to remove, generalize, or redact

- Which captured activity to suppress from the dashboard, and which to

always surface

- Instructions about when related activity should be combined into one entry

They do NOT govern:

- Duration rounding (a setting)

- The auto-merge grouping mode: no grouping / medium / broad (a setting)

- Whether an entry is billable or non-billable (set per client and matter

on the Billing Codes page)

- Task codes and activity codes, including UTBMS (billing code configuration)

- Which client or matter an entry is assigned to as a hard rule (billing

code configuration; guidelines can only assist matching)

- Anything requiring a new client, matter, or code to exist

Take every candidate rule from Parts 2 and 3 and sort it. Anything outside the

governed list above must NOT appear in the paste-ready block, no matter how

correct it is. Move it to the settings or billing-code section instead, and

name exactly where the admin goes to set it.

If a convention needs both — for example, non-billable internal meetings need

a billing code change AND a narrative rule — split it into its two halves and

put each half where it belongs.

───────────────────────────────────

PART 5 — HOW TO WRITE EACH GUIDELINE

Every guideline must tell Billables AI what action to take, not only what to

avoid. Use verbs like remove, exclude, replace, generalize, condense, redact.

Good: "Remove and exclude references to internal staffing decisions, and

generalize the narrative to the legal task performed."

Not good: "Do not include internal discussions."

- One rule per line. If a line contains a semicolon joining two instructions,

split it into two lines.

- Order the rules most important first.

- Never write two rules that could contradict each other.

- Use client and matter names exactly as they appear on our Billing Codes page.

- Do not invent a convention you cannot point to in the sample.

- Never use a placeholder, bracket, or blank to be filled in later. Every line

must be complete and usable exactly as written. If a rule depends on a

decision we have not made, it is not a rule yet — it belongs in the review

section as a question.

═══════════════════════════════════

OUTPUT FORMAT

Produce these four sections in this exact order. The paste-ready block comes

LAST, so it is read in context rather than copied blind.

SECTION A: READ THIS BEFORE YOU PASTE ANYTHING

- Your sample assessment from Part 1, including the timekeeper caveat.

- Open questions: every UNCERTAIN item as a plain question for the firm.

- For each question, the exact rule we should add once we decide, written out

in full with the decision already filled in. If a question has two plausible

answers, write out both versions and label which answer each belongs to.

- Any stylistic pattern you are treating as one person's habit rather than

firm policy, flagged for firm review.

- What you could not determine from this sample, and what additional entries

would help.

SECTION B: BILLING CODE CHANGES (Billing Codes page, not Org Preferences)

Anything involving billable/non-billable status, task or activity codes,

client- or matter-specific rules, or new codes that need to exist. Group by

client or matter name. For each, state what to change and where to click.

If the sample shows none, say so.

SECTION C: SETTINGS CHANGES (Settings page, not Org Preferences)

Anything that is a toggle or dropdown rather than text: rounding, auto-merge

grouping level, narrative separator, whether durations appear inside merged

narratives. For each, give the recommended value, the reason from the sample,

and where in Settings to find it. If the sample shows none, say so.

SECTION D: PASTE THIS INTO ORG PREFERENCES

Start with this single line, then the rules:

(Review Section A before pasting — some decisions are still open.)

Then a continuous list of guidelines, one per line, nothing else. No headers,

no labels, no confidence ratings, no commentary, no explanations, no brackets.

Every line must be one complete instruction that Org Preferences can actually

act on. Nothing from Section B or C may appear here.

═══════════════════════════════════

Here are our example entries:

[PASTE APPROVED TIME ENTRIES — aim for 20 to 25, ideally from several

timekeepers across several practice areas]