A leadership team sits down to talk about GenAI.
The ambition is there. So is the pressure.
Someone has seen what a larger firm is announcing. Someone else has heard a client ask about AI-enabled delivery. There is a sense that the window is narrowing — and that whatever comes next needs to be meaningful, not another pilot that looks good in a presentation and quietly disappears six weeks later. Not another tool, let alone another training.
That is usually the moment when the conversation gets too big too quickly.
What should our GenAI strategy be? How do we compete with firms that have more budget, more data, more people? What is the breakthrough use case that changes everything?
Fair questions. But probably not the first ones worth asking.
James Watt Wasn't Trying to Start a Revolution
James Watt did not invent the steam engine. He was repairing one.
While working on an existing Newcomen engine, he noticed how much steam was being wasted as the cylinder was repeatedly heated and cooled. His breakthrough was not a grand theory of industrial transformation. It was a practical fix to a visible inefficiency: the separate condenser.
That improvement made the engine far more efficient. One insight led to another. Better motion. Better power. Better application. Over time, those compounding improvements helped turn the steam engine into one of the defining technologies of the Industrial Revolution.
The starting point was not a blank page. It was a machine that already existed, a problem that could be observed, and someone paying close enough attention to ask: where is the energy being lost?
Compounding Impact
When firms look at the most visible GenAI leaders in their industry, it is easy to feel like the race has already been run and lost.
But what looks like a sprint from the outside is almost always the visible surface of years of invisible work — pilots, internal experimentation, vendor relationships, data preparation, governance discussions, and the slow accumulation of learning about where the technology actually fits their business.
When Olympians compete, it looks nearly effortless from the stands. That is the point. What you do not see is the decade of early morning practices, failed attempts, and small refinements that made those performances possible.
The right lesson is not to move faster by skipping steps. It is to start compounding sooner.
Don't Make One Big Bet. Make Five Small Ones.
In a dispersed law firm with multiple practice groups, you do not need a single firm-wide GenAI initiative. You are better off running several small, focused ones — in parallel, each owned by the people closest to the work.
Each group focused on what they care most about. One practice group experimenting with knowledge reuse. Another improving how client questions get answered. A third reducing friction in matter setup. Each initiative is modest on its own. Together, they are something more valuable: a rapid learning engine.
If three practice groups are running experiments simultaneously, the firm learns three times faster than if it had placed one central bet. Five groups: five times faster. Perhaps not faster than the competition — but as fast as you can possibly move toward clarity about what works for your firm, in your context, with your clients.
That is the real prize at this stage. Not the tool. The map.
Parallel experimentation only compounds, however, when the lessons are shared. What an associate discovers in one group becomes useful to associates in another. What works in litigation may inform what gets tried next in corporate. The firm builds institutional knowledge that no single initiative could have produced alone.
One important design principle: each initiative needs support and orchestration. That means making resources available — not as an afterthought, but as part of the plan from the start. Underfunded experiments produce noise, not signal. More often, they simply stall.
Where to Look First
Every firm has places where effort leaks away quietly: repeated drafting, knowledge trapped in individual inboxes, client questions answered from scratch, internal tools that people avoid because they interrupt the work instead of supporting it.
Those are not glamorous starting points. But they are where the signal is strongest — if you can see them clearly.
Most firms operate in silos that make this harder than it sounds. Practice groups understand their clients' needs. Pricing knows which groups are underperforming. IT knows which tools are underutilized. But that knowledge rarely travels across those boundaries. Breaking down those silos — even partially — is often the prerequisite to answering the questions that matter most:
- Which clients are underserved and why?
- Which client needs are becoming harder to meet with the current model?
- Where is high-value time being consumed by low-value repetition?
- Where does work slow down because knowledge is hard to find or share?
- What expertise could be reused more effectively across matters or groups?
Start there. Learn what happens. Let that shape the next decision.
The Advantage Is Not in the Tools
A larger firm may have more resources. But a mid-market firm often has advantages that are easy to underestimate: closer client relationships, shorter decision paths, clearer practice focus, and less distance between strategy and the people doing the work.
GenAI does not erase the value of judgment, trust, or specialization. It changes where those qualities can create leverage.
The question is not simply what can this tool do? The better question is what do our clients need — and how could GenAI help us deliver that with more consistency, reach, or speed?
That is where the work becomes specific enough to matter.
The next James Watt is probably not sitting somewhere imagining a revolution from scratch.
They are looking closely at the work already in front of them — and noticing, quietly, where the steam is being wasted.
Getting that clarity is harder than it sounds — and more valuable than most firms expect. For a concrete framework to help you find it, read the business fable: RISE Above AI Chaos.
Rise Above AI Chaos: A Business Fable About Leading Organizations Through the AI Revolution