Article

Law firms: get paid for your records

By TrustPast

Updated November 7, 2026

This page is for the partner of a small firm who has kept every matter for fifteen years and assumed none of it could ever leave the building. Most of it can’t. A law firm holds more of other people’s material than almost any business, and the first job is to separate that material from the firm’s own. What is left is a record of how a practice runs, and buyers license records like that to train AI models.

The thing most firms don’t know they have is not the client files. It is the firm’s own trail of how the work got done: the intake that became an engagement, the budget that moved, the internal email where two lawyers argued over a strategy and settled it. Our page on what data buyers want from a trade business explains the test buyers apply, using trades as the example. Here is the same test applied to a law firm, with the hard limits first in mind.

Every sentence below about the buyer quotes Polyshares’ published pages. Everything about a firm’s files is our own description. Counsel, not this page, decides what your engagement letters allow.

The records a law firm keeps

These are our examples, not a buyer’s list. They are the records a small firm produces as it works, year after year, whether or not anyone thinks of them as a record.

The intake and the conflict check. A call comes in. Someone takes the facts, runs the names against the conflicts database, and writes a short note: clear, a possible conflict, or a decline. Over a decade, the intake notes show which kinds of matters the firm took, which it turned away, and why. The decline note is the interesting one. It records a judgment call made in ten minutes on the phone.

The engagement letter and the budget. The scope of the work, the rate, the retainer, the estimate of hours by phase. Then the revised budget, when discovery ran long or the other side brought a new claim. A firm that kept its original budgets next to the final bills has a record of how legal work is scoped and how far reality moved from the scope.

The time entries. Most firms treat them as billing data. Read the narratives instead. “Review opposing counsel’s response; call with client re settlement range; revise draft letter.” Ten thousand of those entries, by matter type and by phase, show how the work is actually sequenced. The billing system holds the hours. The narrative holds the reasoning in a few words.

The matter workflow. The checklist for a residential closing. The task list for an estate administration. The calendar of deadlines on a litigation matter, with the ones that moved and the reason they moved. The deal checklist with forty items, each marked open, waived or done. This is the firm’s method, and it is mostly the firm’s own. Our page on project history explains why a record of how projects ran, start to finish, is worth more than the finished output.

The drafts and the redlines. The first draft of a lease, the associate’s version, the partner’s markup, the version that went out, the other side’s response. Draft by draft, a document shows how language was argued over. This is also where client material and the firm’s own work are most tangled together, so it is the part counsel reads first.

The internal email. Two partners disagreeing about whether to file a motion. An associate asking how to handle a difficult opposing counsel, and a senior lawyer answering in four paragraphs. The office manager working out why a filing fee was charged twice. Our page on email archives covers what a mailbox holds and how its size is counted. In a law firm, the internal threads are often where the reasoning sits, and they are also where client facts turn up in passing. Both things are true at once.

The firm’s own operations. The knowledge bank of forms and precedents the firm wrote itself. The training notes for new paralegals. The file-retention schedule. The vendor disputes. The marketing pipeline of referrals and how each matter came in. None of that belongs to a client.

How a matter became a matter, how a budget was renegotiated, how a deadline crisis was handled: that is the how. Who the client was is not the point, and it is the part that comes out.

See your estimate

What counts, in the buyer’s own words

Polyshares does not publish a page about law firms. Here is what its pages do say, and how it applies.

On who qualifies, Polyshares’ home page says: “Any business with real operating history, in any sector. Software, logistics, manufacturing, professional services, healthcare administration, financial services, retail, hospitality, construction and more.” A law firm is a professional services business. The same answer adds: “The length and texture of the record matter more than headcount or revenue.”

On what it means by operational data, Polyshares’ data licensing page says: “Operational data means the working record of the business. Slack and Teams, Drive and SharePoint, email, CRM history, support tickets, design files, call recordings, accounting and ERP exports.” A firm’s email server, its document management system and its practice management software map onto that list.

On what it pays for, the same page: “The value is not in volume. It is in reasoning that was captured while the work happened.” And: “A thread where two people disagree about a quote and work it out carries judgment that cannot be manufactured. The messy working record is the point, not a defect in it.” In a firm, the equivalent is the internal thread where a strategy was argued out, or the budget that was revised three times with a reason each time.

On what it does not cover, the home page says the licence covers “Defined material for a defined term. It does not cover what you hold on behalf of your own clients.” That line matters more to a law firm than to any other business on this site. Polyshares’ answer page on whether selling company data is legal says a company holds two kinds of material that “sit side by side and look the same.” One is its own operating record. The other is “material held on behalf of a client, usually under a master agreement that assigns ownership to the client. That second kind cannot be licensed regardless of value, and only the contracts separate the two.”

Polyshares’ SharePoint guide describes what that system usually holds: “SharePoint sites are where teams keep controlled documents: policies, procedures, project files, contracts and quality records.” It adds that “Version history shows how a document moved from draft to approved.” For a firm, the policies, procedures and precedent bank are its own. The client contracts sitting on the next site are not.

What is never included

Start here, because for a law firm it decides everything else.

Client names, contact details, account numbers and anything that identifies a person are on Polyshares’ list of fields its PII partner removes. On its data handling page: “Names, emails, phone numbers, account numbers, and addresses are removed or replaced by a specialist PII partner before Polyshares or any buyer sees the record.” Payment details, trust account records and client billing statements are not how the work was done and stay out, as our own advice. Our page on what is never included sets out the general lines.

Client material is the larger question. A firm holds its clients’ documents, its clients’ secrets and work done for clients under engagement letters. Much of that is the client’s, whatever the firm’s copy looks like. Our page on client-owned work and NDAs explains how the contracts decide what is yours.

Then privilege. The Federal Rules of Evidence, published by the United States Courts, define the terms in Rule 502(g). “Attorney-client privilege” means “the protection that applicable law provides for confidential attorney-client communications.” “Work-product protection” means “the protection that applicable law provides for tangible material (or its intangible equivalent) prepared in anticipation of litigation or for trial.” Those definitions point to “applicable law” for the rules themselves. We don’t interpret any of it. Ask counsel, ideally ethics counsel, before a single count goes on an intake form.

Polyshares says the same on its own page: “This is a description of how these deals are structured and it is not legal advice,” and “a company should have its own counsel review any license before signing.” Our page on whether you need a lawyer goes through when that review happens. Our page on health, payment and other regulated data covers the other categories that carry their own rules, such as a personal injury file with medical records in it.

There is a practical tool here too. Polyshares’ Outlook guide says: “You pick which mailboxes go into the eDiscovery search, so legal, HR or board mailboxes never need to be included.” Its SharePoint guide: “You choose which sites go into the eDiscovery search, so HR, legal and executive sites can be excluded.” A firm can keep every client matter workspace out of scope and still have the firm’s own operations sites, precedent bank and internal administration in it. Whether that leaves enough to be worth licensing is the buyer’s call, after counsel has drawn the line.

How de-identification works

Here is the process Polyshares publishes, in order.

Nothing moves first. Polyshares’ data handling page: “Intake asks for record counts, not files or exports. Nothing leaves your systems until Polyshares has made an offer and you have accepted it.” A firm can count mailboxes and matters without opening any of them.

Then the PII partner. “A specialist PII partner anonymizes the record first.” The fields it names are removed or replaced, and “Roles, timestamps, and the work itself are kept.” For a firm, “roles” means partner, associate, paralegal, opposing counsel. The time entry keeps its sequence and its narrative. It loses the client.

Then the review. “Someone on the team reads the anonymized sample before Polyshares prices the record. The review is scoped to what would actually be licensed.” If anything still reads as identifying, Polyshares says, “the record goes back for another pass before pricing continues.”

Then the listing. “Listings are published without the company’s name, website, people, or exact locations.” And “Two named people are the only ones at Polyshares who see a company’s identity alongside its data. Buyers see de-identified listings only.”

One caution of our own for law. De-identification removes names. It doesn’t change who owns a document, and it doesn’t answer what a firm’s confidentiality duties allow. Those questions come before the PII partner, not after. Our page on how de-identification works covers the general process.

What decides the price

Polyshares prices each company after reading the record. It publishes a range and four deals, and it says what moves the number.

The range: Polyshares’ home page says “Typical deals run $100K to $2M and above” (indicative, not an offer). Its answer page on what data is worth says deals “typically land between $100K and $2M paid to the company” (indicative, not an offer).

The published deals: Polyshares lists four, each a “Perpetual exclusive license to the operational record. Anonymized before delivery.” None is a law firm. The one in professional services is a global marketing agency at $110,000 (indicative, not an offer). We name it only because it is the closest sector Polyshares has published, not as a guide to what a firm would receive.

What raises it, in Polyshares’ words: “Length of continuous operating history, because a record that spans years shows how judgment changed as conditions did. Visible disagreement and correction, because a record of people being wrong and then getting it right cannot be manufactured. A specialized domain, because material from a trade with few written records is scarce in a way general business correspondence is not. And clean rights, because provenance that does not have to be litigated is worth paying up for.”

That last phrase is the one for a law firm. Clean rights are harder to show in a firm than in a roofing company. Everything counsel can confirm as the firm’s own raises the value of what is left.

What lowers it, in Polyshares’ words: “A short operating history is one.” Then “Records the company holds but does not own are another, since material governed by a client master agreement belongs to the client regardless of value.” Then “Personal information woven through the record in a way that cannot be cleanly stripped is a third.” A firm should expect the buyer to look hard at all three.

Polyshares also says “The only way to get a real number is to have a buyer look at the record.” Our page on what decides the price goes through the factors in general. Fifteen years of matter workflows and internal threads, with rights counsel has confirmed, is a different record from three years of a shared inbox. Which is yours?

Which programme fits

On this site, Polyshares’ programme is for US companies with fewer than 30 people. Most small firms, solo practices and boutiques fall inside that line. Our page on the programme for 29 or fewer people sets it out.

The estimate asks how many people work at the firm, counting everyone: partners, associates, paralegals and staff. It names the programme that fits that headcount. If your firm is larger, or based outside the US, the estimate will say so and point you to the right programme.

Polyshares itself says “The length and texture of the record matter more than headcount or revenue.” Headcount decides which programme you apply to. The record decides the price.

Accounting and tax firms face the same split between their own working record and what they hold for clients, with tax return information that carries its own rules; see accounting and tax firms. Imaging and diagnostics centers work under privacy law too, with the scheduling and equipment side of the business usually their own; see imaging and diagnostics centers. Design and web agencies have the same question of client-owned deliverables against the agency’s own process; see design and web agencies.

TrustPast

Questions and answers

Can a law firm license anything at all, given client confidentiality?

Sometimes, and only after counsel has read the engagement letters. Polyshares says its licence does not cover what you hold on behalf of your own clients, and that only the contracts separate the two. The firm’s own operating record, de-identified, is the part that can be in scope.

Does Polyshares see client names or matter details?

Polyshares says names, emails, phone numbers, account numbers and addresses are removed or replaced by a specialist PII partner before Polyshares or any buyer sees the record. It also says roles, timestamps and the work itself are kept, which is why a firm decides with counsel what goes in scope first.

Can we leave the litigation mailboxes out?

Yes. Polyshares’ Outlook guide says you pick which mailboxes go into the eDiscovery search, so legal, HR or board mailboxes never need to be included. Its SharePoint guide says the same of sites.

Do we have to send files to find out?

No. Polyshares says intake asks for record counts, not files or exports, and that nothing leaves your systems until Polyshares has made an offer and you have accepted it.

What does Polyshares pay?

Polyshares says typical deals run $100K to $2M and above (indicative, not an offer). It says it looks at the record before it prices it, so no figure predicts an offer for a particular firm.

What does it cost the firm?

Nothing. Polyshares says there is no fee and no commission, and that money moves one way, from Polyshares to the company. Polyshares pays TrustPast a referral fee; you pay us nothing.