Phase and Task Code Billing, Explained for Firms New to It

The first time a firm gets a set of outside counsel billing guidelines from a corporate client or an insurance carrier, the section that causes the most confusion is almost never the rate table. It is the two or three paragraphs saying every line item must carry a task code and an activity code, that block billing will be rejected, and that the firm is expected to report against an agreed budget by phase. Partners read that and hear administrative overhead. Associates read it and hear a new field to fill in. Both readings miss what is actually happening, which is that the client has asked your firm to describe its work in a shared vocabulary so that the work becomes comparable, auditable and, importantly, budgetable.

Phase and task code billing is not a formatting requirement bolted onto an invoice at the end of the month. It is a decision about how time gets recorded in the first place, made at the moment the attorney logs the entry, by the person who did the work. Firms that treat it as an export problem discover this the hard way, usually around day three of a paralegal trying to retroactively assign codes to six weeks of entries written by four different people who each described a deposition prep session in their own words.

This post is about the mechanics. What the codes are, why the clients who require them genuinely need them, how coding changes an attorney's daily habits, what the codes let you see in your own reporting that you cannot see today, and how the whole thing lands in the electronic invoice at the end of the cycle.

What phase, task and activity codes actually are

The system most US corporate clients and carriers use is the Uniform Task-Based Management System, usually shortened to UTBMS, developed jointly by the American Bar Association, the corporate counsel association and an accounting firm in the mid nineties, originally for litigation and later extended to other kinds of work. It works on two axes at once. The task code says what part of the matter the work belongs to, and the activity code says what you physically did. A litigation task code like L330 sits inside the L300 discovery family and means depositions, while an activity code like A101 means plan and prepare for. Put them together on one line and the entry reads as preparation for a deposition, which is a very different thing from attending one, even though both are three hours of an associate's day.

The word phase refers to the family rather than the individual code. L100 is case assessment, development and administration. L200 is pre trial pleadings and motions. L300 is discovery. L400 is trial preparation and trial. L500 is appeal. Every specific task code lives inside one of those phases, which is why a well coded matter can be sliced by phase without anyone having to tag the phase separately. There are parallel code sets for non litigation work, including counselling, bankruptcy, patent and trademark matters, plus a separate expense code family for disbursements like transcripts, filing fees and online research. The activity codes are shared across all of them, which is the part that trips people up, because A103 for drafting means the same thing whether you are drafting a motion or a licence agreement.

Why corporate and insurance clients insist on them

A general counsel's office or a claims department is not managing your matter. It is managing a portfolio of matters spread across many firms, and the only honest way to answer a question like "what does discovery cost us on an average employment case" is to have every firm describe discovery the same way. Without codes, that question turns into a manual read of thousands of narrative lines written in the private idiom of a hundred different attorneys. With codes, it is a query. That is the entire motivation, and it is a reasonable one. The client is not doubting your integrity, it is trying to run legal spend the way it runs every other category of spend.

Insurance carriers push harder on this than most corporate clients, because their economics depend on it. A carrier defending thousands of claims needs to know whether a given panel firm is spending more on motion practice than its peers, whether a claim's defence costs are tracking against reserves, and whether trial preparation started earlier than the case warranted. None of that is visible in a stack of PDFs. It is visible the moment the same work is coded consistently across the panel. Firms that resist coding tend to find their volume quietly drifting toward the firms that do not, not because anyone made a decision about quality, but because the coded firms are simply easier to evaluate.

The UK runs on a different set, and other jurisdictions differ again

If your firm works with insurers or corporates in England and Wales, UTBMS is not automatically the answer. The costs budgeting regime that came out of the civil procedure reforms uses a phase structure of its own, and the LEDES Oversight Committee published a UK code set, commonly called J codes, designed to line up with the budget phases the court actually looks at, including pre action work, statements of case, case management, disclosure, witness statements, expert reports, trial preparation and trial itself. The point of that alignment is that the same coded time can feed both the client's invoice and the costs budget filed with the court, instead of being reconstructed twice by different people.

Adoption of that UK set has been uneven, and plenty of English firms still bill their insurer clients in UTBMS because the carrier's e-billing platform expects it. Canadian and Australian firms working with international corporates and carriers typically encounter UTBMS through the client's platform rather than through any local rule. The practical instruction is the same everywhere. Do not assume the code set from your last client applies to your next one, and do not assume a code taxonomy carries any regulatory weight on its own. Coding requirements come from the client's guidelines and, in some jurisdictions, from costs rules, and both vary. Confirm which set applies before you configure anything.

What coding does to the way an attorney logs time

Here is the part that firms underestimate. Coded billing is not a change to the invoice, it is a change to the time entry, and the time entry is a habit. An attorney who has spent fifteen years writing "reviewed file and prepared for hearing, 2.4" is being asked to do three things differently at once. Decide which phase of the matter the work belongs to. Decide which specific task inside that phase it maps to. Then write a narrative that supports both, in language a stranger at the client's office can verify without calling you. That is a real cognitive load the first month, and it is much heavier if the entry is written on Friday afternoon about work done on Tuesday.

The firms that get through it fastest are the ones that move time capture closer to the work rather than trying to improve their memory. When time is logged against the matter as the work happens, the code is obvious, because the attorney is still inside the task. When time is reconstructed from a calendar three days later, the code becomes a guess, and guessed codes are exactly what a client's audit software is built to catch. This is why coded billing tends to expose a firm's existing time discipline rather than create a new problem. The firms that were already logging contemporaneously find coding mildly annoying. The firms that were reconstructing weeks in arrears find it close to impossible until they fix the underlying habit first.

Block billing is the habit that has to die first

Almost every set of outside counsel guidelines prohibits block billing, meaning a single entry that lumps several distinct activities into one time value. Coding makes that prohibition enforceable rather than aspirational, because an entry with two activities in it genuinely cannot carry one honest activity code. If the four hours covers a client call, document review and drafting, there is no single code that describes it, and whichever one the attorney picks is inaccurate. This is why block billing and coding cannot coexist. Adopting codes forces the split whether the firm intended that consequence or not.

Splitting entries feels like it will inflate the administrative burden, and in the first weeks it does. What actually happens over a full billing cycle is closer to the opposite. Split entries are easier to defend, so fewer of them get written off during the client's review, and the attorney stops having to reconstruct what a four hour block contained when a reviewer queries it two months later. The firms I have watched go through this properly usually find that their realisation improves not because they billed more time, but because less of the time they billed came back contested. Line level detail is what turns a billing conversation from a negotiation into a verification.

Who owns the code on a time entry, and when

The wrong answer, which is also the most common one, is that a billing coordinator adds the codes at invoice time. That approach guarantees three things. The coder is not the person who did the work, so the code is inferred from a narrative rather than known. The coding happens under end of month time pressure. And any error found later has to be traced back to an attorney who has moved on to other matters. Firms that run this way tend to normalise a small amount of quiet miscoding, which is fine right up until a client's audit reveals a pattern.

The right answer is that the person logging the time picks the code, and the system makes that pick cheap. That means the matter already knows which client requires codes and which code set applies, so the attorney is choosing from a short relevant list rather than searching a full taxonomy. It means the billing type sits on the matter, so an hourly corporate defence matter behaves differently from the flat fee work down the hall without anyone remembering a rule. In Casely, billing type is a property of the matter itself, and hourly, flat fee, contingency and blended arrangements are all native, which matters here because a firm rarely runs one billing model across every client and the coded matters cannot be allowed to dictate how the uncoded ones are handled.

What the codes let you see in your own reporting

This is the part nobody sells the firm on, and it is the part that pays. Once your time carries phase and task codes, your own matter reporting stops being a single total and becomes a shape. You can see that discovery consumed sixty percent of the fees on a case that settled before any dispositive motion was filed. You can see which associates spend disproportionate time on document review versus drafting. You can compare the cost of taking a deposition across every matter your firm handled last year rather than relying on the impression of whoever happened to run them. None of that requires the client. The codes were added for the client, and the firm gets the analytics as a side effect.

That shape is also what makes flat fee and blended arrangements survivable. A firm quoting a fixed fee on employment defence work without coded history is quoting from memory, and memory systematically underestimates the phases that grind, which are usually discovery and trial preparation. A firm with two years of coded matters can price from the actual distribution of where hours went on comparable files. This is the difference between an alternative fee arrangement that is a calculated bet and one that is a hope. Coded time is the raw material for pricing, and firms that have it start winning fixed fee work that firms without it cannot safely quote.

Phase level budgeting is where the codes start paying the firm back

Most sophisticated clients now attach a budget to the matter, and increasingly that budget is expressed by phase rather than as one number. That is a better arrangement for the firm than it sounds, because a phase budget makes overruns visible while there is still time to have a conversation about them. If the discovery phase is at eighty percent of budget with document production not yet complete, that is a call to the client in week six, not a fee dispute in month five. The firms that get burned by budgets are the ones that cannot see phase spend until the invoice is assembled, at which point the only available move is to write time off.

Making this work requires that the matter itself carries structure the firm can report against, not just a pile of hours. A configurable matter stage tracker helps here in a way that is easy to overlook, because the stage a matter is genuinely in and the phase codes being logged against it should agree. When your stage tracker says trial preparation and your time entries are still coded to written discovery, that mismatch is information. Casely lets each firm configure the stage stepper per practice area, so litigation stages and transactional stages do not have to pretend to be the same thing, and the phase codes on the time entries have something real to be checked against.

How coding connects to the LEDES export at the end of the month

The electronic invoice is the destination, not the origin. A LEDES 1998B file expects a task code and an activity code on every fee line and an expense code on every disbursement line, and it has no separate field where a phase can be declared, because the phase is already implied by which family the task code belongs to. That single detail explains most of what goes wrong in practice. If the task code on the entry is wrong, the phase reporting on the client's side is wrong too, and nobody finds out until a spend analysis lands on a general counsel's desk showing your firm doing appellate work on a case that never went up.

The mechanical consequence is that the export can only ever be as good as the coding on the underlying entries, which is why the file should be a formatting step rather than a data entry step. In Casely, unbilled hours on a matter become one itemised draft invoice in a single click, and LEDES 1998B export is supported from that draft, so the codes attached when the time was logged are the codes that travel into the file. Nobody retypes anything, which removes the specific failure mode where a coordinator quietly repairs a missing code by picking a plausible one. If a code is missing, the honest fix is to ask the attorney who did the work, and the workflow should make that the path of least resistance.

Where firms get this wrong in the first ninety days

The most common failure is treating coding as a training problem when it is a systems problem. You can run a lunch session explaining UTBMS to twelve attorneys and still have inconsistent coding a month later, because the constraint is not knowledge, it is friction at the moment of entry. If picking a code takes three clicks and a search, entries will be coded badly. If the matter already knows which code set applies and offers the eight codes actually relevant to this phase of this case, coding takes a second and the accuracy problem largely disappears. Fix the interface before you write the memo.

The second failure is inconsistency between timekeepers, which is really a supervision gap. Two associates on the same matter coding identical work differently will produce a spend profile the client cannot interpret, and the client will notice before you do. Someone senior needs to review coded entries during the first few cycles, not to catch dishonesty but to settle the genuine ambiguities, because a fair number of tasks legitimately sit near a boundary between two codes and the firm needs one answer rather than four. Write those answers down as your house convention, and treat that convention as a living document you revisit when a new client's guidelines say something different.

Start coding before a client makes you

The firms that handle coded billing well almost never started because a client demanded it. They started because someone looked at a matter that lost money and could not explain which part of it went wrong. Coding is the cheapest way to make a matter legible to the people running the firm, and the fact that corporate and insurance clients also want it is a fortunate alignment rather than the main reason to do it. If you wait until a carrier's guidelines arrive, you will be building the habit and meeting a deadline at the same time, which is the worst possible sequence.

Start with one practice area and one code set. Turn on coded entry for new matters only, so nobody is retrofitting history. Give it a full billing cycle before you judge it, because the first two weeks always feel worse than the steady state. Then look at what your own reporting can suddenly answer that it could not before, and decide from there whether to extend it across the firm. The infrastructure question is simply whether your system captures the code at entry, keeps it attached through billing, and carries it into the export without anyone retyping it, which is exactly what legal e-billing software is supposed to do, working alongside the time tracking that feeds it.

Casely is cloud native with nothing to install and a free plan to start at zero, so testing coded entry on a single matter costs your firm an afternoon rather than a procurement cycle. Whatever system you use, judge it on one question. When an attorney finishes a deposition prep session at seven in the evening, how many seconds does it take to record that hour with the right task code, the right activity code and a narrative a stranger could verify? Everything else in coded billing follows from the answer.