Setting and defending a matter budget without guessing

A litigation associate quotes a client £45,000 for a contract dispute. Eight months later the matter has billed £71,000 and the partner is asking why nobody flagged it. This happens constantly, and it's almost never because the fee earner was bad at their job. It's because nobody built the budget on real data, and nobody checked it against actuals until it was too late.

This lesson covers both halves of that problem: building a defensible phase-based budget, and running the monthly checks that catch a matter running hot while there's still time to act.

Where the Numbers Actually Live

Every budget starts in the firm's practice management system (PMS), the software that tracks time entries, disbursements, billing, and matter status. Common platforms include Intapp, Elite 3E, Aderant, and Clio (more common in smaller US firms). These systems record:

• Time entries: hours logged by fee earner, by task, at their billing rate
• WIP (work in progress): unbilled time and costs sitting on a matter
• Realization rate: what percentage of billed value is actually collected, distinct from what percentage of worked time gets billed at all
• Disbursements: third-party costs like court fees, expert witnesses, e-discovery vendors

A budget without PMS data behind it is a guess dressed up as a number. The fix is pulling actual hours and fees from closed matters that resemble the new one.

Building the Phase-Based Budget

Law firms increasingly budget by phase and task code, a structure popularized by the UTBMS system (Uniform Task-Based Management System), originally built for litigation billing but now used across practice areas. Instead of one lump estimate, you break the matter into stages:

Example: a mid-size commercial dispute

The "precedent avg. hours" column comes from pulling 5 to 10 closed matters of similar size, jurisdiction, and complexity out of the PMS and averaging phase-level hours. This is the single biggest upgrade over guessing: you are anchoring the estimate to what actually happened last time, not to what feels reasonable today.

Blended rate is a weighted average of the rates of everyone likely to touch the file (partner, associate, paralegal), not just the lead fee earner's rate. If a partner at £650/hour will only touch 10% of the work, the blended rate should reflect that, not sit at £650 across the board.

Building in Contingency

Precedent-based estimates still miss surprises: an extra expert report, a jurisdictional challenge, an uncooperative opposing party. Firms commonly add a contingency band of 10 to 15% on top of the phase total, stated explicitly to the client rather than buried. A £73,200 estimate might go to the client as "£73,000 to £84,000 depending on scope of disclosure," which is both more honest and more defensible later.

The Monthly Variance Check

A budget that nobody revisits is decorative. The mechanism that actually protects the matter (and the firm's margin) is a monthly variance review: comparing budgeted phase cost against actual WIP for that phase, every month, for the life of the matter.

The variance check has three things to look at:

1. Hours variance. Are actual hours in a phase tracking above the precedent average? If disclosure was budgeted at 60 hours and you're at 55 hours with two more weeks of document review expected, that's a flag before it becomes a crisis.

2. Rate variance. Is a partner doing associate-level work (drafting, not reviewing)? This inflates cost without adding value, and it's invisible unless someone checks who logged what against the phase plan.

3. Phase-timing variance. Has the matter spent longer in a phase than planned, independent of hours? A phase that should take six weeks but is still open at month four usually means scope creep or a stalled opposing party, both of which need a client conversation, not silent absorption of cost.

A Simple Worked Check

Say the disclosure phase was budgeted at £19,200 (60 hours at £320 blended). At month-end, WIP shows 48 hours logged and the phase is 60% complete by task list, not hours.

Run the math: 48 hours logged for 60% completion implies a full-phase forecast of 48 / 0.6 = 80 hours, not 60. At £320/hour that's £25,600, a £6,400 overrun (33% over budget) on that phase alone.

Caught at month-end, this is a conversation: cut scope, renegotiate with the client, or absorb it knowingly. Caught at matter close, it's a write-off or an awkward bill.

Presenting Matter Profitability to the Partner

The monthly figures a partnership actually reviews (typically at a practice group meeting) center on a short list:

• Realization rate: billed value ÷ worked value. US firms average an estimated 85 to 90% realization on hours actually billed at all (as of recent industry surveys like the Thomson Reuters Report on the State of the Legal Market); collection realization (cash collected ÷ billed) tends to run lower.
• WIP aging: unbilled time sitting over 90 or 120 days is a red flag, since aged WIP is disproportionately likely to be written off.
• Matter margin: (fees billed minus direct cost, mainly fee-earner time at cost rate) ÷ fees billed.
• Budget-to-actual variance at matter level, the running total of the monthly checks above.

When you present a matter that ran over budget to the partner who ran it, the instinct is to lead with excuses. Don't. Lead with the number, then the driver, then the fix.

Weak framing: "The client kept adding requests and the other side was difficult, so we went over."

Strong framing: "Matter closed at £81,000 against a £73,000 estimate, a 10% variance, driven almost entirely by disclosure running 20 hours over plan in month 4. We flagged it at the time in the variance review; the client agreed to the extra scope but we didn't formally re-quote. Next time, that triggers a written change order at the point of detection, not at close."

This framing survives partner scrutiny because it shows the control existed and names the specific process gap, rather than blaming the client or opposing counsel after the fact.

Key Takeaways

• Build budgets phase by phase using actual precedent-matter data from the PMS, not intuition; add an explicit 10 to 15% contingency band rather than padding a single number silently.
• A budget only works if paired with monthly variance checks comparing actual hours and phase completion against the plan, catching overruns while there's still room to act.
• Watch three variance types: hours logged vs. plan, rate mix (who's doing the work), and phase timing.
• The metrics partnerships actually track monthly are realization rate, WIP aging, matter margin, and budget-to-actual variance; know these cold before any profitability conversation.
• When presenting an overrun to a partner, lead with the number and the specific driver, then the process fix. Never lead with an excuse.

🎬 [VIDEO: "How Law Firms Track Profitability" — youtube.com — search for practice management and legal finance explainer channels covering WIP, realization, and matter budgeting basics]