Walk into most professional services firms and you will find the same quiet dependency: the business runs on what a few senior people happen to know. The lead attorney knows how to move a case from intake to verdict. The founding consultant knows which engagements will go sideways and how to steer around it. That knowledge is real, valuable, and almost entirely undocumented. It lives in the practitioner’s head, gets passed to juniors through years of apprenticeship, and walks out the door every evening. The firm’s ceiling, whether anyone names it or not, is the personal capacity of the people who hold that knowledge.
The firms that break through that ceiling tend to do one unglamorous thing first. They convert what their best people know into systems other people can run. This is the discipline of codifying expertise, and it is one of the more reliable dividing lines between a practice that stays the size of its founder and one that becomes an institution.
Tacit knowledge is the hidden cap on growth
Economists have a useful pair of terms for this. Tacit knowledge is the kind you can demonstrate but struggle to explain, the feel a surgeon has for tissue or the instinct a negotiator has for when to go quiet. Explicit knowledge is the kind you can write down, hand to someone else, and have them execute without you in the room. Every expert-led business is a mix of the two. The question that decides how far it can scale is what percentage of the core method has been moved from the first column to the second.
When the method stays tacit, growth is linear at best. You can hire more juniors, but each one needs years of proximity to a senior practitioner before they produce at the level the firm is known for. The founder becomes the bottleneck for quality, and often for everything else too. When the method becomes explicit, something different happens. New people ramp faster. Quality stops depending on who is available that week. And the firm can do the thing tacit-knowledge shops almost never manage: absorb a new technology or a new market without losing what made it good.
Here is the diagnostic worth sitting with. Could the core system your firm depends on be taught to a capable outsider in a few days, with a workbook and some follow-up support? If the honest answer is no, your scalability is capped at the personal bandwidth of the people currently running the place, no matter how talented they are.
What codification actually looks like in practice
Codifying expertise is not writing a mission statement or drawing an org chart. It is the harder, more specific work of taking a repeatable outcome and reverse-engineering the steps that produce it, until those steps can be handed to someone else.
In practice it tends to involve a few moves:
- Name the outcome precisely. Not “great client service” but “a case that reaches settlement in under X months with these documented checkpoints along the way.”
- Interview the experts against real cases. The decisions that feel like intuition usually have logic underneath. The work is dragging that logic into the open, one judgment call at a time.
- Write the system down and test it on someone who was not in the room. If a competent newcomer can follow it and get close to the expert’s result, you have codified something. If they can’t, you have written a summary, not a system.
- Keep the version that survives contact with reality. Codified knowledge is not a document you finish; it is one you refine as edge cases show up.
The payoff is that the expertise stops being a person and starts being an asset. Assets can be trained into new hires, extended into new service lines, and, increasingly, wired into software.
One illustrative case
Consider the arc of attorney and entrepreneur Sean Callagy, which is instructive precisely because his constraints forced the discipline early. Callagy was in his mid-twenties, running a growing law firm, when his eyesight began failing due to retinitis pigmentosa. He built that first firm past forty employees and sold it for multiple seven figures while on the verge of losing his sight, then kept building rather than scaling back. His current practice runs well past a hundred people, and he founded a separate medical recovery business that has grown into a large enterprise in its own right. Along the way he earned two Top 100 National Jury Verdicts between 2014 and 2016.
The detail worth extracting is not the résumé. It is that a founder who could not rely on being personally present for everything had strong incentive to make his methods explicit. Callagy describes having codified his approach to human influence into a teachable framework he can transfer to other operators. Whether or not you ever encounter that particular framework, the underlying move is the general lesson: a second firm in a different industry, staffed by different people, is what codified expertise makes possible. It is the same system deployed twice, not two unrelated strokes of luck.
Adversity tends to accelerate codification because it removes the option of doing everything yourself. But you do not need a crisis to reach the same conclusion. You only need to decide that the knowledge in your best people’s heads is too valuable to leave undocumented.
Why this matters more now
Codifying expertise has always paid off in training speed and resilience. Artificial intelligence raises the stakes. AI systems are, at their core, engines for running explicit processes at scale. A firm that has already written down how it qualifies clients, manages a matter, or structures an engagement has something a language model or a workflow tool can actually work with. A firm running on undocumented instinct has nothing to hand the machine.
This is one reason AI adoption moves faster in industries with documentation-heavy cultures, such as software, manufacturing, and parts of finance, and slower in fields like law, where much of the value still lives as individual judgment and decisions run through partner consensus. The lag is not mainly about interest in the technology. It is about whether the underlying work has been made explicit enough to automate. Firms that do the codification work first will find that AI amplifies a system they already trust. Firms that skip it will be automating chaos, or waiting on the sidelines while competitors compound a head start.
The bottom line
If you build, lead, or advise an expert-driven business, the single highest-signal question about its future is not revenue or headcount. It is how much of the firm’s core method has been moved out of people’s heads and into systems. That ratio predicts how fast you can train, how cleanly you can hand off, how well you survive a key person leaving, and how ready you are to put new technology to work.
The practical starting point is modest. Pick the one outcome your firm is best known for. Map the steps an expert takes to produce it, in enough detail that a capable newcomer could follow along. Notice where the map goes fuzzy, because that fuzz is exactly where the tacit knowledge is hiding, and that is where the real value gets unlocked. Then keep refining until the system, and not any single person, is what your clients are actually buying. That is how a founder-led practice finally learns to grow beyond its founder.