One drop, one ripple

Drop a stone into still water and you don't just get a splash, you get a ripple. It moves outward, touches the shoreline, and over time it reshapes the very edge of the water it started in. The next stone dropped into that same water moves differently, because the shoreline it's traveling toward already remembers the last one.

That's the entire argument of this article, compressed into one image.

In a law firm, the stone is a matter. A favorable judgment, a deal that closed against the odds, a piece of regulatory advice that kept a client out of trouble. The ripple is what Business Development does with it: it turns the outcome into a marketing point, a case study, a reason to call a prospective client who fits the same profile. And the shoreline, the thing that remembers, that gets reshaped a little with every ripple that reaches it, is Knowledge Management.

Most firms staff the stone, the ripple, and the shoreline as three separate departments, each with its own head, its own budget, and often its own software. That's not a criticism of any of the three; each stage genuinely requires distinct expertise, and knowledge management in particular is a discipline in its own right, with its own methodology for capture, taxonomy, and governance that most BD teams simply don't have. The argument here isn't that one stage matters more than the others. It's that they are one thought moving through three stages, and firms rarely design the handoffs between them with that in mind.

Every practice throws a different stone

Walk across a firm's practice groups and you'll see the same loop playing out with a different texture each time:

1. Litigation produces a stone shaped like a verdict or a settlement: dramatic, quotable, easy to turn into a headline.
2. M&A and Corporate produces a stone shaped like a closed deal: valuable for league tables, deal-value marketing, and sector credibility.
3. IP and Tech produce a stone shaped like a prosecuted patent or a defended trademark: quieter, but it builds a very specific, very searchable kind of authority.
4. Tax and Regulatory produces a stone shaped like advice that prevented a problem: the hardest to market, because the win is invisible, but often the most valuable to a sophisticated client who knows what near-misses cost.

The stone is different every time. But what happens to it afterward, how it becomes a marketing point, then a reason to target a client segment, then an asset the next associate can find and reuse, is exactly the same mechanical process, regardless of practice. That sameness is the whole case for convergence. If the process downstream of the stone is identical, running it through two separate departments with two separate owners is just friction.

The chain nobody draws on a slide

Here's the chain, made explicit, because it usually stays implicit and that's exactly why firms keep KM and BD apart:

Matter happens, outcome is favorable, BD turns it into a marketing point, the marketing point defines or sharpens a client segment, the segment-specific knowledge becomes a KM asset, the KM asset makes the next marketing point faster to produce.

Notice the loop closes. It isn't a one-way pipeline from BD to KM. The KM asset feeds back into BD's next move. A precedent captured properly doesn't just sit in a database for lawyers to search; it becomes the fastest possible starting point for the next pitch to a similar client. This is why calling KM “downstream” of BD, or a support function to it, undersells what's actually happening. They are the same loop, viewed from two points on its circumference.

What the two disciplines actually share

The instinct in most firms is to keep KM close to the lawyers, under General Counsel, a Chief Knowledge Officer, or Professional Support, and BD close to the client-facing side, under a CMO or CBDO. Historically that made sense, because the two functions were solving genuinely different problems: KM's core discipline is capture and governance, building the taxonomy, maintaining precedent quality, protecting confidentiality boundaries, deciding what counts as a reliable source. BD's core discipline is audience and timing: knowing which client needs to hear what, and when. Neither of those skill sets is a subset of the other, and a firm shouldn't want it to be. A precedent bank run by people without KM's rigor becomes unreliable fast; a pitch strategy run by people without BD's read on the client becomes tone-deaf fast.

What the two functions share isn't a skill set, it's a purpose. Both exist, ultimately, to build brand in the sense that matters to a client: what the market believes the firm is capable of, and how quickly it can prove it. KM builds the proof: a well-governed precedent, a rigorously maintained expertise directory, an honest after-action review. BD carries that proof to the right audience at the right moment. Neither half works without the other; a beautifully governed knowledge base that never reaches a pitch is as much a missed opportunity as a brilliant pitch built on stale or ungoverned knowledge.

That shared purpose is the actual case for structural convergence, not that one function should run the other, but that the handoff between them deserves as much design attention as either function does individually. The numbers back this up: iManage's 2026 Knowledge Work Benchmark, surveying over 3,000 business and technology decision-makers, found that 28% of the most knowledge-mature organizations rank in the top quarter of their industry by financial performance, against just 7% of the least mature, and roughly 80% of the most mature firms operate at a profit, compared with 54% of the least mature. That's not a correlation about knowledge management as a filing exercise. It's a correlation about knowledge management as a growth function, which is exactly the role BD already occupies.

A workable shape:

1. One accountable head over the combined function, chosen for the ability to represent both disciplines credibly: someone who understands KM's governance requirements as well as BD's client instincts, or a leadership pairing where both are represented at the top rather than one reporting several levels below the other.
2. Teams organized by the nature of the work, not by legacy department lines: a pitch/proposal team, a client intelligence and segmentation team, a precedent and expertise-capture team (retaining KM's specialist skill set, not diluting it), an AI/knowledge-systems team.
3. Tier-1 clients handled as a cross-cutting layer above this structure, not a separate vertical, because a Tier-1 relationship draws on all four teams at once, and a merged function should mean that doesn't require four separate approvals to coordinate.

The metrics, and why the intersection earns them

It's easy to list metrics. The useful part is explaining why putting KM inside BD moves each number, rather than just watching it happen and hoping.

Time saved on RFP/pitch turnaround. This is the most direct payoff of closing the loop. When the team writing the pitch is the same team that curated the precedent and tagged the outcome, there's no handoff, no ticket, no waiting for someone in another department to locate the right case study. The evidence and the distribution sit in one workflow instead of two.

Lateral onboarding speed. A lateral hire's biggest cost isn't salary, it's the months spent rebuilding a mental map of “who's done what, and where's the precedent for it.” The stakes here are larger than they might first appear: ALM Intelligence estimates the all-in cost of hiring a lateral partner now averages $2.3 million, and can exceed $5 million for the most sought-after names, while research from Decipher finds that 48% of lateral hires leave their new firm within five years and 62% fail to bring their promised book of business, with a failed lateral at an Am Law 151–200 firm costing an average of $715,000 to replace. An expertise directory built by a merged BD-KM function is built for the exact question that determines whether a lateral sticks: “who's done what, and where's the precedent for it?” A KM team working in isolation tends to build for searchability; a BD-adjacent KM team builds for speed-to-answer, which is what onboarding actually needs.

Cross-sell/referral rate uplift. Cross-selling is fundamentally a knowledge problem disguised as a relationship problem: you can't refer work you don't know a colleague has done. Passle's 2026 “Collaboration Gap” report found that 81% of firms are falling short on cross-selling and growing client accounts through collaboration, and that 58% of firms are instead leaning on rate increases as their primary growth lever, with 54% of those firms admitting that same pricing strategy is the leading cause of client churn. Firms are, in effect, paying a churn tax for a problem that better-connected knowledge would solve more cheaply. When client intelligence and matter/precedent data live in the same function, the system that flags “this client also has a need in X” is drawing on real matter history, not a partner's memory at a conference dinner.

Knowledge reuse rate (search-to-use). This is the metric that most exposes a poorly designed handoff. KM's governance discipline is what makes a precedent trustworthy enough to reuse in the first place; that's not the part to change. What tends to go missing is a feedback loop back from BD on what actually got used in a live pitch and what didn't, so the taxonomy can be tuned toward real demand. Reuse rate goes up when that feedback loop is built into the workflow, rather than left to informal hallway conversations.

AI grounding accuracy / reduction in hallucinated content. This is the newest metric and arguably the one that makes the whole argument urgent rather than theoretical. A 2025 Stanford and Yale study published in the Journal of Empirical Legal Studies tested the leading AI legal research tools and found hallucination rates of 17% for Lexis+ AI and 33% for Westlaw AI-Assisted Research, against 43% for general-purpose GPT-4 — a reminder that even purpose-built legal AI, grounded in licensed case law, still gets a meaningful share of answers wrong. (Worth noting honestly: this kind of research dates quickly, since a study published in a given year is usually testing the previous year's model versions — but the pattern it reveals, that grounding quality is the main lever on accuracy, has held up across successive studies.) AI tools are only as good as the knowledge base they're grounded in, and that quality bar is set almost entirely by KM's governance standards: accuracy, currency, and structured tagging. What convergence adds is making sure that governed knowledge is tagged in a way that reflects how BD and clients actually ask questions, not just how a librarian would file it. Get both halves right, and the same knowledge base that grounds internal research is trustworthy enough to sit behind a client-facing AI tool.

Back to the shoreline

None of these metrics are really about efficiency for its own sake. They're all downstream of the same fact: every ripple that reaches the shore reshapes it, and every stone dropped afterward moves faster because of the shape that's already there. A firm that keeps its stone-throwers and its shoreline-keepers in separate departments is, in effect, choosing to forget the shape of the last ripple every time. A firm that merges them is choosing to remember, and remembering, compounded matter after matter, is what a brand actually is.

Sources

1. iManage, 2026 Knowledge Work Benchmark (survey of 3,000+ business and technology decision-makers)
2. Passle, The Collaboration Gap: Cross-Selling and Collaboration Report 2026
3. ALM Intelligence, lateral partner hiring cost estimates
4. Decipher, lateral partner failure-rate research
5. Magesh, Surani, Dahl, Suzgun, Manning & Ho, “Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools