Built inside a law firm: the Judge Law story, six months on

Most software for law firms is designed somewhere else — a product team, a pitch deck, a set of assumptions about how firms work. Kefilex wasn't. It was built by Kefilab, an independent software company, embedded inside Judge Law — a working UK firm — and it has now been running there live for six months. This is what that actually looked like: the situation, the gap, and — the useful part — what six months of live running genuinely found.

The situation

Judge Law was doing the things a firm is supposed to do. Ads running. A website that ranks. A visible profile. Enquiries coming in every day, by phone, by form, by email, through referrers.

And then, after the enquiry arrived: nothing you could see. Not nothing happening — people were answering phones, replying to emails, opening matters. But no view of it. Nobody could say how many enquiries came in last month, which source they came from, which ones went quiet, or what any of it converted into.

That's not a criticism of the firm. It's the normal state of a UK law firm, because most firms are run by solicitors, not operators. The people in charge are excellent at the law and busy doing it. Nobody's job is to stand over the pipeline and ask where every enquiry went — so nobody does, and the information quietly doesn't exist.

The gap

The obvious answer is "the practice management system handles that." It doesn't.

Clio — which Judge Law runs on, and which Kefilex sits on top of — is excellent at what it does: matters, time recording, billing. Once a matter exists, Clio manages it properly. What it doesn't give leadership is the one connected view: pipeline into matters into money. What came in, what it became, what it's worth, and where the leaks are between those three.

We went looking for the tool that did that for a UK firm on Clio. It didn't exist. Plenty of CRMs that don't understand legal practice; plenty of legal tools that stop at the matter. Nothing that connected the whole line.

So we built it inside the firm. Not as a prototype with sample data — as the system Judge Law actually ran on, with the firm's real enquiries, real matters and real ledger, replacing one manual workflow at a time.

The stack, for the record — five systems that had never shared a number, all feeding one pipeline once Kefilex sat on top (what connects, and how →):

Six months of live running — what it actually found

This is the part a pitch deck can't fake. When you run a real firm's data through a system that connects pipeline, matters and money, things surface. And to be clear before the list: none of what follows is unusual, and none of it is neglect. These are the things the standard toolset hides in every firm — we know, because we went looking for the tool that would show them and it didn't exist. The difference at Judge Law is that the firm chose to look.

Over half the "leads" weren't leads

More than half of the firm's inbound contact wasn't prospective clients at all — it was the Land Registry, the courts, recruiters, suppliers. Count those as "enquiries" — which every firm using a shared inbox does, because nothing separates them — and every conversion statistic gets quietly distorted. Kefilex now filters these out automatically, so the pipeline counts prospects and nothing else.

The biggest lead source was invisible

Phone calls were the firm's largest source of new business — and completely absent from reporting, because a phone call doesn't log itself. Once calls were captured and automatically tagged by practice area, something stranger surfaced: the firm's largest practice area had been missing from lead reports entirely. The busiest part of the firm generated the least visible pipeline, purely because its clients pick up the phone.

The "outstanding" headline nobody should trust

Every firm has an aged-debt number, and in almost every firm it's added up from reports that don't talk to each other. The first time Kefilex connected the ledger to the matters behind it, nearly half of that headline turned out not to be genuinely chaseable at all — some already covered, some already paid, some relating to bills that had never actually been sent. That isn't carelessness; it's arithmetic done across disconnected systems, which is how the standard toolset works everywhere. The difference is what happens once you can see it: an honest chase list half the size, and the credit-control conversation finally about the right amounts with the right clients.

The balances no report has a name for

Beyond the chase list sit the in-between states every practice accumulates invisibly — a payment received but not yet matched to its bill, a balance still resting on a matter that has finished. Nothing lost, nothing wrong, just nothing moving, because no single screen shows them. Kefilex turns them into a named, dated resolution list that gets worked through as routine housekeeping — instead of being discovered in bulk years later, which is the industry norm.

Follow-up runs on rails now

The firm's follow-up used to live in people's heads: someone meant to call that enquiry back, someone was sure a colleague was handling it. Now an enquiry that goes quiet gets chased automatically, and a past client who comes back is recognised as one — routed differently to a brand-new lead, with their history attached, instead of being greeted as a stranger.

Referrers can see for themselves

Judge Law's broker referrers used to phone for updates on the matters they'd introduced — which cost the firm time and the brokers patience. They now log into a portal and see a live milestone timeline for their own referred matters. Nobody has to ask, and nobody has to answer.

What this proves

None of the above was predicted in a planning document. Every one of those findings came out of a real firm's live data, and every feature in Kefilex exists because Judge Law needed it that week — the non-lead filter because the stats were wrong, call tagging because the biggest source was invisible, the credit-control view because the outstanding number wasn't honest, the referrer portal because brokers kept phoning.

That's the argument for building inside a firm rather than alongside one. The problems weren't imagined and then validated; they were found, in the order a working firm actually hits them. Six months on, Judge Law runs on the result — and the product going to market is the one a real firm already depends on, not one a real firm is yet to try.

One thing worth being plain about, because careful buyers rightly ask: Kefilex is owned and operated by Kefilab, an independent software company. Judge Law is its first customer — with no stake in the product and, like every firm on the platform, no visibility of any other firm's data. The proving ground was a law firm; the vendor is not.