If your firm's "knowledge management" strategy relies on one person's memory and a hope that Sarah is in the office, I'm afraid you don't really have knowledge management. You have a collection of smart people covering for a system that doesn't exist yet.
And the cost of that gap? It doesn't announce itself. There's no alarming line item in the P&L. Instead, it trickles out as rework, write-offs, slower matters and that low-level, ever-present anxiety of have I missed something? - a question no fee earner should need to ask as often as they do.
Across conversations with UK legal services firms, one theme keeps surfacing. The knowledge is there. It's just not findable when it matters. And when knowledge isn't findable, people do the most rational thing available to them: they redo work, they re-read, they ask around, they over-prepare. They build a private safety net. It feels responsible. It's also quietly expensive.
Before we go further, one important clarification. Omnient - which I'll come to later - doesn't replace your case management system. It works alongside it. This guide is about the whole picture: the tools, the people and the habits that form when everyone is busy and the stakes are high. No theory. No hand-wringing speeches about how KM is important. Just what's actually going wrong, what it costs and what good looks like in 2026.
This is a hypothetical scenario, but it's stitched together from patterns we hear a lot - so if it feels familiar, that's rather the point.
It's Tuesday. 4:47pm. A client calls wanting a straight update before they brief their board. The fee earner is good - calm, professional, knows their stuff. But the story of the matter is scattered across half a dozen systems. So they do what everyone does:
They search Outlook for the last "final" attachment. They skim a Teams thread. They open the case management system for the official notes. They check the DMS because the document might live there instead. They go back to email because the key decision was buried in a reply chain three weeks ago.
Twenty minutes goes. Not because anyone is slow. Because the firm's knowledge is split across too many places, none of which talk to each other particularly well.
Here's the thing: that twenty minutes doesn't just "waste time". It changes behaviour. Getting up to speed takes too long, so people stop trusting search and start asking the team instead. Which means interruptions, delays, duplicated effort and a client who can sense the wobble - even if you don't say a word.
That's what knowledge management is really about. Not a library. Not a tidy folder structure. It's whether the firm can answer the questions that matter, fast, with confidence.
Most firms talk about KM like it's a storage problem. A bank of precedents. A know-how folder. An intranet page that looks respectable for about a month before it gathers dust.
That's not useless. But it's not the point either.
Knowledge management in a legal services firm is closer to this: Can a new starter find the right context without interrogating three colleagues? Can someone covering a matter understand what's happened and why in minutes, not hours? Can you answer a client question without reconstructing the timeline from scratch? And can the firm see what it already knows before somebody creates yet another version of the same document?
If the honest answer to any of those is yes, but only if Sarah is in the office, then you've built a people-dependent system. And people-dependent systems don't scale. They also don't survive change - holidays, resignations, lateral hires, maternity leave, a partner moving teams. Knowledge lives in people, which means continuity depends entirely on availability. The result is slowdowns, rework and a firm that feels permanently "busy" even when it's properly staffed.
This is why KM has become a leadership conversation, not a back-office one. It's not about tidying up SharePoint. It's about whether your firm can function when the people who hold the answers aren't in the room.
Let's take the blame out of it. Silos don't form because people are selfish or lazy. They form because the work is complex, time is tight and nobody wants to be the person who missed something.
So people create safety. They save the "important" email in a personal folder. They keep a private OneNote with the stuff they know they'll need again. They message the colleague they trust rather than wrestling with the search function for the third time that day. They build a local stash of things I know are good.
It's entirely rational behaviour. It's also how firms drift into a peculiar reality where there are five versions of "the truth" and everyone is slightly anxious about which one actually matters.
There's a systems reason too. Many firms run a case management system, a document management system and a suite of communication tools that don't naturally tell a single story. The Law Society has pointed out that disparate systems which don't share data create inefficiencies and problems - and that's putting it politely. The gaps between systems get filled by people, which feels manageable until you do the maths.
And the maths isn't pretty. One UK review found lawyers and paralegals can lose around 2.3 hours a week searching for - but not finding - the right documents, and another two hours recreating documents that couldn't be located. An IDC-backed study estimated this contributes to roughly a 9.8% loss in productivity per lawyer, per year. That's not a rounding error. That's nearly one day in ten spent on work that shouldn't need doing.
The pain spreads beyond fee earners, too. Operations feels it as capacity loss and process drift. IT feels it as a constant stream of "can we improve search?" tickets and shadow tool adoption. Partners feel it as slower matters and clients who start chasing updates they shouldn't need to chase.
So here's a question worth sitting with: if your systems were truly working, why does the firm still rely on "who knows" to get anything done?
Yes, you do. And you need it. But a case management system is not the same thing as a firm's working memory. It captures what it's designed to capture - and then real life fills the gaps around it.
The "real story" of a matter still happens in email threads, meeting notes, informal Teams conversations and all the unofficial places where context actually lives. I'd call this a split brain: official record on one side, lived reality on the other. The CMS contains the formal record, not the full picture. Which means people still have to hunt for why something happened and what changed - and that hunt leads to slower decisions and the kind of mistakes that come from missing detail.
So let's be clear about what you're solving here. You're not trying to replace your CMS. You're trying to stop people paying a daily tax to reconstruct the context that sits around it. Later in this guide, I'll explain how an "answers layer" fits into that picture.
It's tempting to file this under "nice efficiency gain" and move on. But I'd go further than that - it's a profit leak, and often a significant one.
Legal Futures, reporting on BigHand research, has highlighted how firms are grappling with write-off pressures, including the finding that 59% of firms confirmed write-offs had increased over the past year.
Now, knowledge silos aren't the only cause of write-offs - I'm not suggesting a single tool will fix your profitability overnight. But they are one of the few causes you can actually reduce without asking people to work longer hours. Because a lot of that write-off pressure stems from work that feels necessary in the moment but doesn't translate into clean, billable output. The re-reading, the double-checking, the internal detective work - it all takes time, and much of it never makes it onto an invoice.
A quick note on wellbeing too, because it's part of the environment you're operating in. LawCare's Life in the Law 2025 report found that 59.1% of respondents scored in the "poor mental wellbeing" range, with 78.7% saying they regularly work beyond their contracted hours. One tool doesn't "solve wellbeing" - that would be a daft claim to make. But if your firm keeps funding unnecessary friction, it shouldn't come as a shock when people feel stretched.
Most firms don't need a bigger knowledge base. They need knowledge that behaves differently. Here's what "good" tends to look like in practice:
A useful way to think about it: KM isn't where you store. It's how you retrieve and act. If retrieval is slow, KM is slow. If retrieval is unreliable, KM is unreliable. And if KM is unreliable, people don't change their behaviour. They just work harder. Which, circling back to the wellbeing data above, is already a problem.
This is the practical change we're seeing. Firms are moving away from "everything must be put in one place" thinking. Not because it's a bad idea in principle - it's just a very hard promise to keep in a live firm, with live work and clients who don't wait for you to finish your information architecture project.
Instead, the direction is simpler and more honest: keep your core systems, but add a layer that helps people get to the right context, fast. That layer needs to do a few things well. It needs to search across the places knowledge already lives, rather than demanding you move everything first. It needs to bring back relevant answers, not just a list of documents you then have to read through yourself. And it needs to show where those answers came from - because legal work demands confidence, not guesses.
Omnient is that answers layer - sitting across your CMS, DMS and email, not replacing any of them. To be explicit, because clarity matters: Omnient is a knowledge assistant that gives your firm conversational search across its existing systems.
It's built for the "Tuesday 4:47pm" moments described earlier. The moments where you don't need more information - you need the right context, now, with proof of where it came from.
One honest limitation worth stating: no tool removes the need for professional judgement. But tools can remove friction. And friction is where your time goes.
1. Get the 10-minute KM audit by email. If you want a quick way to put numbers on what's currently just a feeling, we've built a short "findability audit". It's designed for Ops and IT to run without turning it into a project. Three checks from it to give you a flavour: how many places does client context live today? How long does "getting back up to speed" take before key calls? And how often do people ask a colleague because search isn't trusted? If those questions make you pause, you'll probably find the full audit useful.
2. Request early access to Omnient (limited onboarding capacity). If you're already past we should probably look at this and you want to see what an answers layer looks like in practice, get in touch for a personalised demo. We're onboarding a small number of firms each month - mainly because doing it properly takes time, and rushing it would rather defeat the point.
Knowledge management fails quietly. Nobody sends round a memo about it. It just shows up as longer matter cycles, more client chasing, more write-offs and teams who feel permanently behind despite working flat out.
You already have the knowledge. Which means the waste is avoidable. And that's perhaps the most frustrating kind of cost there is - the kind you keep paying because it's become normal.
If you're curious, get the audit by email and run it with your Ops and IT teams. You'll know soon enough whether this is a small annoyance or a real profit leak.
My guess? It's probably the latter.
