Law firms have been buying the wrong software for 20 years. The right software was never going to be off-the-shelf. Now the economics have shifted enough that the better-managed firms are figuring this out.
The Off-the-Shelf Failure Pattern
The pattern is consistent across firms of all sizes. A managing partner returns from a legal technology conference convinced that software X will transform the practice. The firm licenses it, spends three months on implementation, and watches adoption stall at 20% because the software doesn't match how attorneys actually work.
Generic practice management software assumes generic legal workflows. But legal workflows are extraordinarily specialised. A mid-size IP litigation firm has case management needs that look nothing like a real estate transactional practice. Both look nothing like a high-volume immigration shop. The taxonomy of document types, deadline calculations, billing workflows, and conflict checks varies dramatically by practice area.
When you force attorneys to adapt their workflows to software, they don't. They use the software for what it does easily and keep doing everything else in email and spreadsheets. The result is two systems of record, neither of which is authoritative.
What Changed: The Cost of Custom Dropped
Custom software used to mean: hire a development firm, spend 18 months, pay £500K, get something that barely works and nobody will maintain. The model is broken because it treats software as a capital project rather than an evolving system. But the economics of custom development have improved substantially. Modern frameworks and AI-assisted development have compressed certain categories of work—particularly CRUD interfaces and document processing pipelines—to a fraction of what they cost two years ago.
A document intake and routing system that would have taken four months to build in 2022 takes six weeks today. A conflict check integration with a firm's matter management system that once required a dedicated integration consultant now requires a well-documented API and a competent engineer.
The ROI Is Computable
The firms investing in custom software are doing so because they can compute the return. Attorney time is expensive—billing rates at AmLaw 200 firms average £800-1,200 per hour for partners, £400-600 for associates. Any workflow that eliminates 30 minutes of manual data entry per day for a team of 10 associates is worth £750K-£1.1M annually in recovered billable time.
Matter intake automation. Document review routing. Client portal workflows that eliminate phone tag. Deadline calendar automation that pulls from court filing systems. These are not glamorous projects, but they have measurable, computable ROI.
The critical change is that firms are now asking their operations teams to document time spent on administrative workflows and presenting that data to decision-makers alongside software proposals. When the operations team can show that associates spend 2.3 hours per week on tasks that could be automated, the investment decision becomes straightforward.
The Data Sovereignty Concern
One reason law firms resisted cloud software for years was client confidentiality. Putting client matter data on a SaaS vendor's shared infrastructure felt risky—not necessarily because it was, but because it was difficult to represent to clients and in engagement letters.
Custom software on dedicated infrastructure solves this cleanly. Firms can run their applications in isolated cloud environments with defined data residency, audit logging that meets their specific requirements, and access controls that map to their actual permission structures.
What Good Implementation Looks Like
The firms that succeed with custom software share a few practices. First, they start with a specific, well-defined problem—not "we need better case management" but "associates spend 40 minutes per day entering billing data that already exists in our email system." The narrower the initial problem, the higher the likelihood of a successful first project.
Second, they involve attorneys in design, not just requirements gathering. Attorneys have strong opinions about workflow that they often can't articulate in the abstract but can recognise immediately when they see it. Building a clickable prototype before writing production code saves months.
Third, they treat the first project as a proof of concept for the broader relationship between the firm and its software capabilities—not as a one-time engagement. The firms extracting the most value from custom software have an ongoing development relationship, shipping improvements on a monthly cadence, adjusting to feedback, and expanding the scope of automation over time.
The technology investment that pays off in legaltech isn't the biggest or most sophisticated. It's the one that solves a real, expensive problem for the people who have to live with it every day.