Legal Billing Software Development: Trust Accounting, LEDES E-Billing and What It Costs

Legal billing software development costs $30,000 to $450,000. A time-capture and invoicing MVP runs $30,000 to $55,000, a small-firm system with trust accounting runs $55,000 to $110,000, and a mid-size platform with LEDES e-billing runs $110,000 to $220,000.

Two features separate attorney billing software from every other billing product, and both are where projects go over budget. The first is trust accounting, because client money has to sit in a separate account and reconcile three ways. The second is LEDES e-billing, because corporate and insurance clients will not pay an invoice that arrives in the wrong file format.

Generic billing software handles neither. That is the whole reason law firms commission custom builds, and it is the part most cost guides skip.

Legal billing software development cost at a glance

Why attorney billing is not ordinary billing

A standard invoicing system assumes you bill for a thing at a price. Legal billing breaks that assumption in five places, and each one is a build cost.

Client money is not firm money: Retainers and settlement funds belong to the client until earned. They sit in a trust account, and touching them wrongly is an ethics violation rather than an accounting error.

One matter can have several fee arrangements: Hourly for the litigation, flat fee for the filing, contingency on the outcome, with a blended rate for junior associates and a split between two originating partners. The rate engine has to resolve all of it on one invoice.

Partners edit bills before they go out: The pre-bill review step, where a partner rewrites narratives, writes off time and adjusts entries before the client sees anything, has no equivalent in normal billing software. It is a workflow, with versioning and approvals.

Corporate clients dictate the invoice format: They require LEDES files with UTBMS task codes, and their e-billing portals reject anything malformed.

Time is captured, not entered: Lawyers bill in six-minute increments across dozens of matters a day. If capture is awkward, time goes unrecorded and the firm loses revenue the software was bought to protect.

Firms that need billing for a non-legal business are usually better served by custom billing platforms built for standard invoicing, since none of the above applies.

Trust accounting: the most underestimated module in the build

Trust accounting is the single most expensive module in legal billing software, typically $14,000 to $32,000 and 6 to 10 weeks. Teams that have not built it before usually budget a third of that.

Under the ABA’s Model Rule 1.15 on safekeeping property, a lawyer must hold client funds in a separate trust account, keep complete records of those funds, and retain the records for a period set by the governing state rule. Advance fees go into trust and come out only as they are earned. A lawyer may deposit personal funds in a trust account only to cover bank charges, and only up to the amount needed.

Translated into software requirements, that means:

• Separate ledgers per client and per matter, never a pooled balance
• Three-way reconciliation between the bank statement, the sum of client ledger balances, and the trust account ledger, which has to tie exactly
• A hard block on overdrawing any client ledger, even when the pooled account has money in it
• Earned-fee transfers that move money from trust to operating only against a specific invoice, with an audit record
• No negative balances, ever, and no netting one client against another
• An immutable audit trail, since trust records are what a bar examiner reviews

The requirement that breaks naive implementations is the third one. A pooled IOLTA account holds money for many clients at once. A system that validates against the account balance rather than the individual client ledger will happily let one client’s matter spend another client’s retainer. That is commingling, and it is the kind of defect that ends careers rather than causing a support ticket.

Rules vary by state, so the reconciliation and retention logic has to be configurable rather than hard-coded.

LEDES e-billing: why invoices get rejected

If the firm bills corporate clients or insurance defense work, LEDES export is mandatory, not optional. Budget $10,000 to $24,000 and 4 to 8 weeks, plus ongoing maintenance.

LEDES is the Legal Electronic Data Exchange Standard, the invoice file format corporations use to receive bills from law firms. Per the LEDES 1998B specification, the format is pipe-delimited with 24 fields, carries fields such as INVOICE_NUMBER, CLIENT_ID, LAW_FIRM_MATTER_ID and LAW_FIRM_ID, and can hold several invoices in one file. UTBMS task and expense codes travel in the LINE_ITEM_TASK_CODE and LINE_ITEM_EXPENSE_CODE fields when the client requires them.

The part that costs money is not writing to the exporter. It is that every client’s e-billing portal validates differently. One requires task codes on every line, another rejects them on trademark matters, a third caps narrative length, a fourth refuses invoices where block-billed entries exceed a time threshold. A rejected invoice means the firm waits another billing cycle to get paid.

So the realistic build is an exporter plus a validation layer that checks a draft invoice against each client’s specific billing guidelines before anything is submitted. Firms that skip the validation layer spend the first six months after launch chasing rejections manually.

Beyond 1998B, newer LEDES XML versions exist and some clients require them, so plan for more than one output format.

Cost breakdown by module

Firms that also want general ledger and payroll inside the same system should scope accounting software development alongside this rather than stretching the billing product to cover it.

Developer rates by region

For this particular build, prior trust accounting experience is worth more than the rate difference. A team that has shipped three-way reconciliation before will get it right in six weeks. A team learning it will spend ten weeks and still need a legal accountant to review the output.

Build custom or buy Clio, MyCase or PracticePanther

Buy for small firms. Build for mid-size and larger, or where fee arrangements and e-billing requirements do not fit what the products support.

Practice management products with billing typically run $40 to $140 per user per month depending on tier. Modelling at an illustrative $100 per user per month:

Check current vendor pricing before using these figures in a business case, since tiers change.

Price is not the only input. Build when the firm has fee structures the products cannot express, when e-billing guideline validation across many corporate clients is a daily cost, when the firm wants billing inside an existing system, or when the practice has an unusual area with its own billing conventions. Buy when the firm is small, the work is standard, and the real requirement is getting off spreadsheets this quarter.

The in-between option that often wins: buy the practice management product, build only the e-billing validation and reporting layer the product does not do well, and connect the two.

What it costs to run after launch

• Maintenance and updates: 15% to 20% of build cost a year
• Hosting: $400 to $2,500 a month depending on firm size and document volume
• LEDES and guideline maintenance: budget for it annually, since corporate clients revise billing guidelines and portals change validation rules
• Annual security review: $3,000 to $10,000, and worth doing given what sits in a matter record
• Payment processing: per-transaction, and note that surcharging client payments is restricted in some jurisdictions

Where legal billing projects overrun

From delivery patterns rather than theory:

Trust accounting is scoped as “an accounts module”. Covered above. It is the most common cause of a blown timeline on these builds.

Rate rules discovered late: Firms describe billing as hourly, then produce a matter with a blended associate rate, a partner discount, a fee cap and a split origination. Get ten real historical invoices during discovery and price the engine against those, not against the description.

Pre-bill review underestimated: Partners want to edit narratives, write off time, move entries between matters and see the version history. It is a workflow with approvals, not a screen.

LEDES rejections after launch: Without the guideline validation layer, the firm finds out about format problems from the client’s portal, one invoice at a time.

Migration from legacy practice management: Historical time entries, trust balances and matter records rarely come out clean. Trust balances in particular have to migrate exactly, since the opening reconciliation must tie.

How Comfygen approaches these builds

Comfygen has delivered software since 2019, with 550+ projects for 400+ clients across 30+ countries and 97% client retention, under ISO 9001:2015, ISO/IEC 20000-1:2018 and ISO 27001.

On a legal billing build specifically:

• Discovery starts with ten real historical invoices, because the rate engine has to handle what the firm actually bills rather than how it describes billing
• Trust accounting is built first and reviewed by an accountant familiar with the firm’s state rules before the rest of the system goes on top
• Three-way reconciliation is tested against a seeded dataset with deliberate mismatches, not just a happy path
• LEDES export ships with a validation layer against each client’s billing guidelines, in the first release
• Document and matter data is access-controlled from day one, since privilege is not something to retrofit

Firms that want reporting on realization, utilisation and matter profitability alongside the billing system usually scope that as a business intelligence layer reading from it, rather than building dashboards into the billing product.

Conclusion

Legal billing software is priced by two modules, not by a feature count. Trust accounting and LEDES e-billing decide the budget, the timeline and whether the system survives a bar audit or a client’s billing portal.

Everything else on the list, time capture, invoicing, reporting, payments, is ordinary software that any competent team can estimate. Get those two right and the rest follows. Get them wrong and the firm ends up reconciling trust balances in a spreadsheet, which is what it was trying to stop doing.

Before commissioning a build, pull ten real invoices and the current trust reconciliation. Those two documents tell a development team more than any requirements brief.

Frequently asked questions