Ten Years of Matter History Is Not a Pricing Asset. Yet.

By Richard Burcher, Global Vice President, Pricing & Profitability at Aderant

Ask a managing partner whether the firm has good data for pricing and most will say yes. A decade or more of matters sit in the practice management system.

Then ask someone to pull the last five matters that look like the one you’re pricing today. This is where the emails start. Someone opens a spreadsheet. A senior partner says what he remembers it cost, and everyone nods because nobody can prove otherwise.

Firms have plenty of volume. What they lack is comparability, and forward-looking pricing runs on comparability.

What “comparable” means

Start with the numbers, but that’s where most firms stop.

Staffing comes first. A matter run mostly by partners prices differently from one run by associates, and a comparison that ignores seniority will mislead you in either direction. Timing matters too. Rates, staffing and efficiency all drift, so last year’s matter is a better guide than one from five years ago, even if the older one looks closer on subject. Then there’s what was charged. The figure that counts is realized fees, meaning what was billed and collected. The rate card only records what the firm hoped to earn.

And how often do partners clone a two-year-old matter without any regard to the two rate reviews that have occurred in the intervening period?

That’s the quantitative half, and the firm can at least see it in the system. But that doesn’t get you to a comparable matter.

Next is the shape of the matter itself. Practice area and matter type are the obvious ones. Jurisdiction and forum change how a dispute runs. So do the amount at stake, the number of parties and the volume of documents. How it ended matters as well. A matter that settled at mediation and one that went to trial are different animals, even if they started out identical.

Then there’s the human layer, which is where the real variance sits. Take two matters that are identical on paper: same type, same forum, same value, same team. In the first, the firm across the table is commercial and pragmatic. They narrow the issues, answer correspondence and pick up the phone. In the second, the other side is pedantic and obstructive. Every point is contested and every request is answered late. Nothing in the engagement letter separates the two

matters, yet the bills come out vastly different.

Put both in front of someone pricing a new matter and the spreadsheet says they’re twins. A partner who ran them would say they have nothing in common. The same applies to clients. Some give clear instructions and decide quickly. Others reopen scope every fortnight. The judge and the deadline change the cost as well.

This is why experienced partners distrust a raw comparison, and they’re right to. Comparability is a judgment about quantity and quality together. A system that matches only on the numbers will hand you a “comparable” that isn’t one.

The numbers were never where the entire story lived. The practice management system holds hours, rates and invoices. What made a matter easy or hard sat in the firm’s own records and in the memory of whoever ran it, who may have left since.

Virtual Pricing Director (VPD) reads those records. Go back to the two matters above. When VPD surfaces them, you don’t see two identical totals with different outcomes. You see that one ran smoothly because the other side was pragmatic and the second was hard going because the other side was obstructive. The person pricing the new matter can then decide which of those it’s likely to resemble, and price accordingly. That’s the difference between a match that looks right on a spreadsheet

and one a partner will trust.

Why the data won’t cooperate

Matter history lives in a lot of places. Billing systems hold some of it. The DMS holds some, and so do spreadsheets with names like “v7_final”. A good portion exists only in partners’ heads. Time narratives are free text, written in a hurry by people who were never asked to write for a pricing model. Matter codes differ between offices and between eras. Phase and task tagging is patchy where it exists at all. Write-offs happen without explanation. Rate codes from a system migration six years ago no longer map to anything.

So, you can usually say what the last matter cost. You can’t say with any confidence what the next one will or should.

This matters more now than it did two years ago, because firms are putting AI on top of this data. A model fed inconsistent inputs will give you confident, inconsistent outputs. That is worse than having no automation, because the answer arrives looking authoritative. Whatever AI-assisted scoping you build is only as good as the structure underneath it.

Two ways to fix it

There are, broadly, two schools of thought.

The first says build your own taxonomy. Design a custom one for classifying matter and work data, then map every practice area to it and keep it maintained. It can be done, but the tool has nothing useful to say until the taxonomy exists. That means months of work before anyone sees output, followed by permanent upkeep. The firm’s data also ends up reinterpreted through a classification model somebody invented for the occasion.

The second anchors to a standard the industry already uses. We use UTBMS. Billing teams know it. E-billing platforms are built around it. You enrich the firm’s own history against that standard instead of designing a new schema, and the firm’s data stays the source of truth. The software organizes it. It doesn’t replace it.

The usual objection is that a standard is too blunt for the way a particular firm works. Fair, and it’s why the mapping runs both ways. Each UTBMS phase maps to the phase the firm actually uses, so L300 Discovery might show up as Depositions & Document Review, and people keep working in their own language. Matters created fresh in VPD and matters from ten years ago run through the same mapping, so you end up with one comparable data set instead of two that don’t talk to each other.

Speed matters as well. A standard that someone else maintains costs you less to keep current than a schema you have to curate yourself.

What the work involves

There are four stages, and none of them is glamorous.

You bring the data in, and not as a one-off migration. A migration is stale the day it finishes. The feed needs to sync continuously, with errors tracked record by record so nothing gets orphaned or quietly dropped.

Then you normalize it. Free-text narratives and metadata become structured fields. Historical fees get resolved against the rate hierarchy that applied at the time, not today’s. Legacy rate codes get assisted remapping instead of being left to rot.

Next comes categorization. Time entries and matter phases are tagged, including on older matters, and then overlaid with the firm’s own taxonomy.

Only then do you put the data to work. Matter context from the PMS, DMS, and CRM goes into a single enriched payload: practice area, jurisdiction, how many depositions, trial or settlement, hours and fees by phase and task. When AI helps scope a matter, it’s reasoning over that history instead of a blank prompt.

What you get for it

Four things follow from a clean data layer, and they build on each other.

Matter matching improves, because the system can compare on structure and on meaning, then rank past matters by a relevance score. Scoping improves, because you can set phase and task budgets from the hours and fees of matters that really are comparable. Tracking gets sharper, since budget and actuals sit in the same standard phases, so an overrun shows up where it started. Margin monitoring follows, rolling up from phase to matter to client.

See it in practice

Register today for Beyond the Spreadsheet: What It Actually Takes to Make Historical Matter Data Pricing-Ready, a Legal Value Network webinar sponsored by Aderant, on Tuesday, November 3, from 12:00- 1:00 pm CT. The session draws on work across 300+ firms and includes a live look at how matter-matching and relevance scoring work in practice. Registration is complimentary for members and non-members.