A Case Pipeline Is Not a Sales Pipeline
Most CRMs model deals that close. Matters don't close, they run for years, and the difference breaks every stage-based tool a firm tries to use.
A firm buys a well-known CRM, spends a fortnight configuring it, and abandons it inside six months. This happens constantly, and the usual diagnosis — bad adoption, partners resistant to technology — is almost always wrong.
The tool was built to model a deal. A deal has stages, moves forward, and ends. A legal matter does none of those reliably, and the mismatch shows up in ways that make the software feel actively wrong to use.
Where the model breaks
Matters go backwards. A case in negotiation returns to discovery. A settlement falls through and litigation resumes. Sales pipelines assume forward motion and treat regression as an error or a lost deal.
Time is the wrong axis. A sales pipeline gets nervous about a deal open for ninety days. A personal injury case is supposed to be open for eighteen months while the client finishes treatment. Every ageing report screams at you until you stop reading them.
There is no "close". A matter resolves, then has a fee disbursement, then a lien negotiation, then a file retention obligation measured in years. "Won" is not the end of the record.
Two pipelines share one object. Before signing, you have a lead: qualify, conflict check, retainer, sign. After signing, you have a matter: entirely different stages, different urgency, different people. Cramming both into one pipeline means half your stages are irrelevant to half your records.
Value is not knowable. Sales pipelines are built to forecast off deal value. A contingency matter has no value until it does. Weighted-pipeline forecasting produces confident numbers that mean nothing.
What a case pipeline actually needs
Two connected pipelines. Intake, with a real end point — signed or declined, with a reason. Then matters, which begin where intake ends. They share a client record but not a stage list. Keeping them separate is the single biggest structural improvement most firms can make.
Stages that describe status, not progress. "Waiting on medicals", "in treatment", "demand sent", "in negotiation". Some are dead time by design, and the system should not treat dead time as failure.
Deadlines as first-class objects. This is the real difference. A sales CRM has a close date. A matter has a statute of limitations, filing deadlines, response windows, court dates — each with its own consequence for being missed. They need to be dated, owned, alerted on well in advance, and impossible to lose in a note field.
Time-in-stage against a per-stage expectation. Not one global ageing rule. Thirty days in "waiting on medicals" is normal; thirty days in "retainer sent" is a problem. The system should know the difference.
Conflict checking at the front. Before a lead becomes a matter, the check has to be prompted, recorded and dated. Bolting this onto a sales tool as a checkbox is how firms end up with an audit trail that does not exist.
What to report on instead of forecast
Sales reporting asks what will close this quarter. Case reporting should ask different questions:
- Matters by stage, with time-in-stage against what is normal for that stage
- Upcoming deadlines, by owner, on a rolling horizon
- Matters with no client contact past their interval
- Intake conversion by source, which is the one genuinely sales-shaped number a firm has
- Fee realisation by matter type, once matters resolve — the closest thing to a forecast that is honest
None of that comes out of the box in a deal-oriented tool, which is why firms end up exporting to spreadsheets and eventually stop using the CRM at all.
The practical question
Practice management software exists and covers a lot of this. For many firms it is the right answer and this article is an argument for buying one rather than building.
Where a custom build earns its place is when the firm's process does not fit the packaged assumptions — an unusual referral structure, a volume intake operation with its own qualification logic, several practice areas that genuinely work differently, or an existing system that has to stay. The test is not whether you could build it. It is whether what you need is meaningfully different from what you can buy.
The honest caveat
Plenty of firms run well on a shared calendar, a good paralegal and a diary system that has worked for twenty years. Software does not make a badly run firm well run, and a pipeline nobody updates is worse than a whiteboard everybody looks at.
The specific failure worth avoiding is the one at the top: buying a sales tool, forcing matters into it, and concluding from the wreckage that your firm is not a technology firm. It was never the firm. It was the model.
We build case pipelines, intake systems and deadline tracking for US law firms — including where an off-the-shelf tool has already failed. See how it fits together, or tell us how your matters run.