The Training Was Great

Finding a recognizable GenAI use case is rarely the hard part.

The real hesitation I see is more fundamental:

Why change something that already works?

Most attorneys:

• Are good at what they do
• Are paid for how they work today
• Don’t feel immediate pressure to do things differently

From their perspective, GenAI isn’t solving a painful problem.
It’s introducing uncertainty into a system that already rewards competence.

This is why early adopters often have a few similar traits: they’re overloaded, stretched thin, or personally feeling the limits of their current workflows. That creates motivation.

But that motivation doesn’t exist for everyone—and adoption efforts that assume it does quietly stall.

What I see working is not trying to manufacture urgency, but reframing what improvement looks like.

Instead of showing what GenAI can do in general, firms that see adoption stick show how it supports and sharpens work that is already good.

Not speed. Quality is the differentiator.

One firm challenged attorneys with before-and-after examples of real work products—emails, outlines, first drafts. The “before” versions were solid. The “after” versions were better in ways attorneys immediately recognized.

One partner’s reaction captured it perfectly:

“I could do this myself. But this gets me to a better starting point than I usually have.”

That’s motivation—not because something is broken, but because there’s visible leverage.

For most attorneys, the real risk isn’t inefficiency.
It’s reputational.

Motivation shows up when trying GenAI feels professionally safe, and when the upside is framed as augmenting expertise rather than replacing it.

That’s when the question quietly shifts from “Why would I change?” to “Why wouldn’t I?”

Here’s what quietly kills adoption:

An attorney sees a polished use case from someone with years of experience.
They try it themselves.
The output is… mediocre.

No one warned them this was normal.

They don’t know whether:

• They did something wrong
• The tool doesn’t work
• Or GenAI is simply overhyped

So they stop.

The firms where behavior change sticks do something different. They set expectations upfront:

“GenAI is predictive, not deterministic. Think weather forecast, not calculator. That difference takes getting used to. The first few attempts will feel clunky—and that’s normal.”

Then they give people one small, safe place to start.

Not “use it for everything.”
But “try it for this one task you already do several times a week.”

Low risk.
Familiar context.
No reputational downside.

Confidence builds.
Overwhelm fades.

Even motivated, capable attorneys forget to use new tools once they’re heads-down in work.

I hear this all the time:

“I believe GenAI could help. I just forget about it when I’m busy.”

The firms where usage sticks don’t solve this with more training.
They solve it with triggers.

Examples that work:

• Making GenAI the first tab people open in the morning
• Defining cue moments:
  • Blank page → ask for structure
  • Too long → summarize
  • Unclear thinking → outline

One attorney described the shift this way:

“I don’t decide whether to use it anymore. In certain moments, it’s just what I do.”

That’s what a good trigger does.

When legal tech usage fades after training, one ingredient is almost always missing:

• Missing motivation:
  Attorneys couldn’t see how the tool improved work they already do well.
• Missing ability:
  The learning curve felt too steep, unpredictable, or unsafe.
• Missing trigger:
  Nothing reminded people to use the tool when it mattered.

The fix depends on which ingredient is missing—not on running more training by default.

Your training team isn’t failing.
They’re being asked to create behavior change using tools designed for knowledge transfer.

The firms seeing traction aren’t replacing training.
They’re surrounding it with:

• Motivation tied to professional judgment and quality
• Safe ability-building with clear first steps
• Simple triggers that fit into existing routines

That’s when training becomes the starting point for change, not the endpoint.