Matter Budgeting That Clients Actually Trust

Almost every fee estimate that blows up follows the same shape. Someone at intake asks how much the matter will cost, a partner does a rough mental calculation based on the last case that felt similar, and a number gets said out loud. That number then goes into an engagement letter, gets forgotten by everyone at the firm, and resurfaces eight months later when the client opens an invoice that is well past it. At that point the conversation is not about the work. It is about whether the firm was honest at the start, and no amount of well written narrative on the bill fixes that.

The uncomfortable part is that the number was usually not wrong in a lazy way. The partner who quoted it had genuinely seen matters like this one. What went wrong is that the estimate was a single figure covering a process with six or seven distinct stages, each of which behaves differently, and there was no mechanism anywhere in the firm for noticing when one of those stages started running hot. A budget you cannot report against during the matter is not a budget. It is a guess with a dollar sign on it.

This piece is about the version that works. Build the budget at the phase level from time data your firm already has, report actuals against it while the matter is still moving, and disclose variance early enough that the client can still make a decision about it. Firms that do this do not have fewer overruns. They have overruns that clients accept, which is an entirely different business outcome.

A budget is a forecast with a structure, not a promise with a number

The first thing to fix is the vocabulary, because it shapes what the client hears. An estimate stated as one figure sounds like a commitment, and clients hear it as a ceiling regardless of how many qualifying sentences follow it. A budget stated as a set of phases with a range on each one sounds like what it actually is, which is a forecast built on stated assumptions. The moment you present it as a structure rather than a number, the conversation moves from whether you will hit it to which assumptions are most likely to move, and that is a conversation a sophisticated client will happily have with you.

The structure also does something quieter and more valuable inside the firm. A single number cannot be wrong until the very end, which means nobody can act on it. A phase budget can be wrong in week three, on one line, in a way that is visible and specific. That is the entire point. You are not building a document for the client file. You are building an early warning system that happens to be shareable, and the phase boundaries are where the warnings fire.

Your own time entries are the only honest budgeting data you have

Industry averages are close to useless for this. Every firm staffs differently, drafts differently, and has a different tolerance for how much partner time goes into a first draft, and a benchmark from a survey of firms you have never met will not tell you what your associates actually take to get through document review. The data that predicts your next matter is the time your own people logged on the last fifteen matters of the same type. That data exists in your system already. Almost nobody looks at it before quoting.

This is the argument for taking time entry discipline seriously that has nothing to do with billing. Every entry that gets logged the same day with a real description becomes a data point you can use to price the next matter. Every entry that gets reconstructed from memory at the end of the month, or never logged at all, is a hole in your forecasting. Firms that treat time capture as a revenue issue are only seeing half of it. It is also the raw material for every fee conversation you will have for the next three years, which is why the practical answer is legal time tracking software that makes same day logging faster than avoiding it.

Phases, not a single number at the top of the page

A phase is a chunk of work with a natural start, a natural end, and a decision point at the boundary. For litigation the phases are close to universal in shape even where the terminology differs, running roughly from case assessment and pre action work, through pleadings, disclosure or discovery, witness and expert evidence, pre trial preparation, trial itself, and any post judgment or appeal work. For a transaction the phases are due diligence, structuring, drafting, negotiation, conditions and completion, and post completion filings. For a family matter the phases follow the procedural milestones rather than the calendar. The specific list matters far less than the fact that there is one.

There is an existing vocabulary here worth borrowing rather than reinventing. The UTBMS phase and task code sets, which corporate clients and their e-billing platforms already use, give you a ready made phase structure and the codes travel inside a LEDES 1998B export. In England and Wales, costs budgeting in multi track civil claims runs through a formal court process on a prescribed form with phases already defined for you, so the structure is imposed rather than chosen. Other jurisdictions handle this very differently or not at all, so confirm what your own court rules and your regulator actually require before you assume any of this applies to you.

Building the first phase budget from history you already have

Start with one practice area and one matter type. Pull your closed matters of that type from the last two or three years, and for each one, group the logged hours by phase. If your time entries carry a phase or task code, this is a report. If they do not, the first pass is manual and it is still worth doing, because you only have to do the archaeology once before the structure starts maintaining itself. What you want out of this exercise is not an average. Averages hide the thing you need. You want the spread, the median, and the specific matters that ran long, along with the reason each of those ran long.

That last column is the whole exercise. When you look at the three discovery phases that consumed double the median hours and you find that all three involved a client whose documents arrived unindexed, you have not found an estimating error. You have found an assumption. Your budget for the next matter of that type can now say, in writing, that the disclosure phase assumes documents are provided in a reviewable format, and that if they are not, that phase moves. Casely's contact labels and connected matters make this history easier to pull than it sounds, because related matters link to each other with the reason stated, so the pattern across a family of similar cases is visible instead of buried across seven separate files.

The assumptions are the budget, and they belong in writing

A range without assumptions is just a wider guess. What makes a phase budget defensible is the short list of conditions it depends on, written in plain language at the point the client agrees to it. The volume of documents you expect. The number of custodians. Whether there will be one expert or three. How many rounds of mark up you have priced for. Whether the other side has a reputation for interlocutory applications. Each of these is a lever that moves a specific phase, and naming it converts a future argument into a pre agreed trigger.

Put those assumptions where they survive staff turnover, which means attached to the matter itself rather than living in the engagement letter attachment nobody reopens. In Casely the matter carries its own record, the stage tracker shows where the work actually is, and the documents attached to it each carry a comment field recording what changed and why, so the budget and its assumptions age with the file instead of drifting away from it. When an associate who joined in month four needs to know whether three rounds of drafting was in scope, the answer should be one click away, not a question that has to go up to the partner who ran intake.

Reporting against the budget during the matter, not after it

Here is where most firms fall down, and it is not a knowledge problem. Every firm knows in principle that you should compare actuals to budget. Very few have a rhythm that makes it happen without someone remembering to. The fix is boring and it works. Once a fortnight, or at every phase boundary for faster moving matters, someone runs the numbers on active matters over a threshold value and looks at three things per phase, which are hours logged, percentage of the phase budget consumed, and an honest read on how much of the phase work is actually finished.

That third one is the one that catches problems early, because hours consumed and work completed come apart long before anyone notices. A discovery phase that is at sixty percent of budget and sixty percent complete is fine. A discovery phase at sixty percent of budget and twenty percent complete is a conversation you need to have this week, not in three months. This is where one click invoicing quietly earns its place in a budgeting workflow, because converting unbilled time into an itemised draft at any point is also the fastest way to see exactly what has been consumed against a phase, on a real document, in real numbers, rather than a dashboard estimate.

What a mid matter budget update should actually say

A good update is short, specific, and boring. It states where each phase sits against budget, flags the one or two lines that are moving, gives the reason in a sentence, and states what you propose to do about it. It does not apologise, because there is usually nothing to apologise for, and it does not bury the variance in a paragraph of narrative. If disclosure is running twelve percent over because the other side produced four thousand more documents than the assumption allowed for, say exactly that, and say whether you expect to absorb it, recover it in a later phase, or need a revised figure.

Send it through a channel that creates a record. This is one of the reasons a real client portal beats email for anything budget related, because the update sits with the matter, the client can open it on their phone, and there is no question six months later about whether it was sent or which version was sent. Casely's portal is privilege filtered per document automatically, so a budget update visible to the client sits alongside internal work product that is not, without anyone having to maintain two folders and hope nobody mixes them up.

Early variance disclosure protects the relationship, and late disclosure destroys it

Clients do not actually expect legal budgets to be exact. Anyone who has run a business understands that litigation and negotiation involve another party whose behaviour you do not control. What clients cannot forgive is finding out late, because late disclosure removes their ability to make a decision. A client told in month three that the expert phase is likely to run thirty percent over can decide to narrow the issues, accept the cost, or push harder toward settlement. The same client told in month nine has only one available reaction, which is anger, and they are entitled to it.

There is a version of professional courage involved here that firms underrate. Telling a client early that a phase is running over feels like admitting a mistake, so the natural instinct is to wait and hope the next phase comes in under and washes it out. That instinct is why so many fee disputes exist. The firms that hold onto their best clients through difficult matters are the ones that send the awkward update in week six, and the compounding effect over a decade of matters is not subtle. Early disclosure is not a confession. It is the thing that proves the budget was real in the first place.

Fixed fee and contingency matters need budgets more, not less

There is a persistent belief that budgeting only matters on hourly work, which gets the economics exactly backwards. On an hourly matter an overrun is a client conversation. On a fixed fee matter an overrun comes straight out of the firm's margin, and on a contingency matter it comes out of a recovery that might not happen. The client may not need a phase budget on those matters, but the firm absolutely does, because it is the only mechanism that tells you whether your flat fee pricing is holding up before you have priced forty more matters the same wrong way.

The practical requirement is that your system supports every model natively on the same matter set, so time gets logged on a fixed fee case even though nobody is billing hourly against it. Casely handles hourly, flat fee, contingency and blended billing as first class arrangements rather than workarounds, which means a flat fee matter still accumulates real time data, and that data feeds the next round of pricing. Firms that stop logging time on fixed fee work lose the only evidence they had about whether the fee was right.

Where budgets quietly break, and it is almost never the estimate

Scope creep is rarely a dramatic event. It is a series of small requests, each entirely reasonable on its own, none of which felt big enough to raise at the time. A quick extra memo here, a call with the client's accountant there, a second look at a clause that was already settled. Nobody ever decided to expand the engagement. It expanded anyway, and the phase budget that was accurate at signing is now measuring a different matter than the one being worked.

The defence is procedural rather than heroic. Every request that falls outside the stated assumptions gets logged against the matter with a note saying it was out of scope, in the moment, by whoever received it. That takes about fifteen seconds and it changes the entire character of the eventual conversation, because you are no longer asserting that scope grew, you are showing eleven dated entries where it grew. This is also where deadline and stage tracking pay off, since a matter whose stage tracker has been sitting in the same phase for six weeks past the forecast is telling you something before the numbers do.

Reporting the budget across the whole book, not one matter at a time

Individual matter budgets are useful to the client. Aggregated budget performance is what tells the firm whether its pricing is sound. Once you have phase budgets running on a reasonable number of matters, the question worth asking every quarter is which phases across which matter types consistently run over. That pattern is almost always narrow. It is usually one or two phases in one or two practice areas, and it usually has a structural cause, such as a phase that is systematically under scoped, a staffing mix that puts the wrong seniority on the wrong task, or a client type whose behaviour your assumptions never accounted for.

Fixing a systematic overrun is worth far more than winning any individual fee argument, because it changes every future matter of that type rather than one invoice. It might mean repricing that phase, changing who does the work, or writing a sharper assumption into the engagement terms. Whatever the answer, you cannot find it without the underlying phase data being consistent across matters, which is the real reason to standardise your phase structure early rather than letting each partner invent their own. Reporting across the book only works if the categories are the same book to book.

Start with one matter type and the last twenty files

You do not need a firmwide initiative for this. Pick the matter type you run most often, pull your last twenty closed files of that type, group the hours by phase, and write down the reason for every overrun you find. That exercise takes a focused afternoon and it will change the next fee conversation you have, because for the first time you will be quoting from your own evidence rather than your own memory. Then quote the next matter of that type as phases with ranges and stated assumptions, and put a recurring fortnightly check in the diary against it.

The tooling matters less than the habit, but the tooling determines whether the habit survives a busy month. If pulling phase actuals means exporting to a spreadsheet and rebuilding a pivot table every time, it will happen twice and then stop. If the matter carries its own time, its own stage, its own documents and its own billing arrangement in one place, the check takes minutes and it keeps happening. That is the case for running this on proper matter management software rather than on discipline alone, and Casely is free to start on, so you can test the workflow on one practice area before committing the firm to it.

What clients trust is not accuracy. It is the sense that you are watching the number as carefully as they are, and that they will hear about a problem from you before they hear about it from a bill. A phase budget built on your own history, reported during the matter and disclosed early when it moves, gives them exactly that. It also gives your firm the one thing a single number estimate never could, which is the ability to be wrong early enough to do something about it.