Your champions are no longer experimenting.
They have found useful workflows, built confidence and changed parts of how they work. In some practice groups, GenAI is already helping with research, drafting, preparation, review or client responsiveness.
That matters. It proves the technology can create value inside the firm.
But concentrated success is not the same as scale.
The harder question is how to move from a small group of sophisticated users to meaningful participation across a much broader part of the firm — not every attorney, not every workflow, and not all at once. A more realistic ambition might be participation across half the firm over the next year, or across the three to five practice groups where the opportunity is clearest.
That is still a significant change. And it requires more than asking the champions to keep demonstrating what they've learned.
Champions help prove value. Scaling requires the firm to make participation rational for everyone else.
The Next Group Is Making a Different Decision
The attorneys who adopted first had a reason to lean in — reduced admin work, strategic advantage, curiosity, or simply enjoying being early. They were willing to tolerate ambiguity and invest time before the return was clear.
The next group is different. They aren't opposed to GenAI. They're evaluating it against a different set of demands: client commitments, team development, business development, risk, billable work. Time spent learning a new way of working has to displace something else. For them, the question isn't can GenAI do something interesting — it's is the return worth the disruption. That's a higher bar, and what can look like resistance may just be an unresolved trade-off.
This is usually where firms turn back to the people who created the initial momentum — asking champions to run demonstrations, share workflows, answer questions across practice groups. It's useful. Champions bring credibility and real examples. But it asks them to solve a problem they never faced themselves: they had an internal reason to begin, and were prepared to experiment before the outcome was certain.
The next group may need the outcome to be visible first, but that is probably not enough. They want to know what is in it for them, and how their efforts will pay off. In other words, they are asking about the return on investment (ROI) — for them.
Table Stakes Justify Access. ROI Justifies Change.
There's a growing argument that GenAI is table stakes — that competitive pressure and client expectations make doing nothing increasingly hard to defend, to the point where the ROI question feels almost beside the point. That may be true for investing in the tools.
Unfortunately, table stakes only takes a firm so far. It can justify buying the technology. It doesn't, by itself, justify changing how attorneys work.
Access is an organizational decision. Adoption is a series of individual and practice-level decisions about time, risk, priorities and value. Attorneys still need a reason to change established habits, and practice leaders still need to decide whether a new workflow is worth introducing into how their teams serve clients. That's where ROI becomes essential — not a narrow hours-saved calculation, but the visible evidence that makes participation rational. If the value stays theoretical, generic, or concentrated among a small group of enthusiasts, the rest of the firm has little reason to make room for it.
Table stakes may justify access. ROI justifies change.
ROI Alone Isn't Enough
Demonstrating potential return is necessary. It isn't sufficient. A firm can know a workflow could create value and still fail to embed it — because the return is too distant, the support too thin, the expectations unclear, or the whole thing dependent on the goodwill of a few already-stretched people.
This is where RISE is useful — not as an alternative to ROI, but as a way to test whether the initiative is actually set up to deliver the return it promises. A compelling use case shows that value is possible. Value becomes a return only once outcomes are defined, the investment matches the ambition, teams get support when the work gets difficult, and expectations are clear enough to sustain accountability.
- Returns: Have specific outcomes been defined, and is someone accountable for delivering them?
- Investment: Does what the firm has committed match the scale of what it expects back?
- Support: Is there active help when teams hit technical, organizational or political barriers?
- Expectations: Are review points, responsibilities and measures of success clear?
You can use the RISE Diagnostic to test whether your GenAI initiative is set up to deliver the returns expected of it.
If one of these is missing, the initiative may keep producing impressive examples without changing how the broader firm works.
Returns
The return that motivated the champions may not motivate everyone else. Early adopters may have valued novelty, visibility, or the chance to build a new capability. A practice leader may care more about whether GenAI improves responsiveness without increasing review risk. The broader the promise, the more work the audience has to do to translate it into their own practice. GenAI can improve productivity isn't a reason to change. This can reduce the time spent building the first draft of a recurring client update is much closer.
That specificity also has to exist at the initiative level, not just the individual one — because the two are connected. If nobody can point to what was committed to, and by whom, when the investment was approved, the initiative is unlikely to deliver any results at all.
Investment
Many firms scale adoption by borrowing time from the people who created the initial momentum — champions keep developing their own use while documenting examples, supporting colleagues, and running demonstrations. This can work for a while. It also creates a predictable tension: every hour spent helping the firm scale is an hour off client work, team leadership or business development, with no real change to what the champion is otherwise accountable for.
This is a familiar pattern in any volunteer-driven initiative, and there's little reason to expect GenAI champions to be the exception. Left unaddressed long enough, it tends to show up first as fatigue, and sometimes later as attrition — not always dramatically, but a champion who quietly stops volunteering, or who leaves the firm because what was meant to be one of the more rewarding parts of their practice turned into an unpaid second job. It's a predictable consequence of unfunded enablement, not a rare one, and considerably easier to prevent than to reverse.
If broader participation is strategically important, it needs capacity of its own — protected time, dedicated enablement support, practice-specific workflow development, or formal recognition for adoption work. The specific investment varies. What matters is whether it's proportionate to the ambition. A strategic priority can't scale indefinitely through volunteer effort.
Support
Champions learn by exploring — testing approaches, accepting uneven results, adjusting as they go. The next group may need something different: a starting point based on a familiar matter type, a tested workflow, a sense of what a good first result looks like. Telling them to experiment can sound open and empowering. But it will come across as an undue burden to people who haven't yet decided the effort is worth it.
Support also isn't only about training. Teams hit technical constraints, competing priorities, unclear ownership, or resistance from stakeholders whose cooperation is required — and at that point, another demonstration won't help. Someone needs to remove these barriers. If teams are largely left to figure out the hard parts themselves, the firm may mistake a support problem for a motivation problem.
Expectations
Champions can unintentionally become the benchmark. Their demonstrations look impressive because they reflect months of accumulated experience — and when that becomes the implicit standard, a newcomer's first useful workflow can look insignificant by comparison.
Meaningful progress looks different across the firm: one partner using an approved tool to draft the first structure of a client update, one team leader testing a recurring workflow with a junior and reviewing the result together, one practice group building a repeatable approach around a single high-friction task. These steps look modest next to an advanced champion's work. They're still how adoption gets embedded.
Expectations also need to be clear at the initiative level — when progress gets reviewed, what each group is responsible for, what evidence decides whether the initiative is working, what would trigger a change in approach. The goal isn't to turn every attorney into a GenAI expert. It's to help more attorneys use the technology appropriately, confidently, and where it genuinely helps. Clear expectations are what make progress visible enough to act and sustain momentum.
For most firms, all of this will be unbillable work. As long as those expectations are clear, that's fine. Just a recommendation: start tracking this unbillable work now, so that next year you know not only what you're asking of participants, but whether it's still fair to call it volunteer work.
The Real Scaling Question
The next stage isn't about proving GenAI works. Your champions may have already done that. It's about whether the conditions exist for the value to spread — and whether the next set of initiatives is set up to convert broader participation into a return the firm can actually point to.
Table stakes got you access. RISE is what decides whether that access ever becomes adoption.
Not a checklist. Not a fixed sequence. A way to find out, while a firm is still early enough for it to matter, whether the conditions for success are actually in place.
Use the RISE Diagnostic to see where your GenAI initiative currently stands →