Many leaders operate on a false assumption: if we train someone on it, they'll do it correctly.
We know that isn't true.
Everyone who has run a team has watched a well-trained employee freeze, guess, or do the wrong thing months after training ended.
And yet most operations, especially in financial services and healthcare, are still built as if training equals accuracy. That gap matters more in regulated industries than almost anywhere else.
The procedures most likely to be forgotten are the ones with:
- The most compliance risk attached to them
- The most likely to produce an audit finding
- The most likely to leave a customer feeling like the person helping them doesn't know what they're doing.
This isn't just a minor training inefficiency. It's where the business absorbs its worst risk and its worst customer experiences at the same time.
What Has to Be True for Training to Work
If performance is what actually matters, the question isn't "did we train them." It's "will they internalize this well enough to do it correctly, on their own, under pressure."
That's a much higher bar, and "training" can only clear it if three things are true:
- The process isn't so complex that it's unrealistic to memorize
- The employee does it often enough to build real muscle memory across all its variations
- The process isn't changing constantly. Every change means employees are working off what they used to know
Why Regulated Industries Fail This Test
A teller answering a balance question hits all three. So does a patient access rep verifying a routine insurance eligibility check. Both are simple, both happen constantly, and neither changes much. Traditional training works fine here.
But every regulated business also carries a long list of procedures that fail all three at once.
Examples of banking procedures that fail the training test:
- Wire transfers
- IRAs
- Trust setup
- What to do when an account holder has died
Examples of healthcare procedures that fail the training test:
- Prior authorization appeals
- Coordinating benefits across multiple payers
- Handling a request to amend a record under HIPAA
Complex, infrequent, and constantly shifting as policy changes. No amount of classroom time or role play fixes that. You can't practice your way to muscle memory on something you touch once a month.
And this is exactly the long tail where compliance errors, audit findings, and damaged customer trust come from. Not the frequent, simple stuff. The rare, complex stuff nobody gets enough reps on.
Where the Gap Actually Shows Up
In practice, this shows up as escalations. An employee isn't sure how to handle a procedure, so they call a coworker or a supervisor, not because they don't understand the business, but because they can't hold the whole procedure in their head.
Supervisors experience this as death by a thousand cuts. Constant interruptions, never enough uninterrupted time to do their own job, because they've become the institutional memory for everyone else's edge cases.
We've seen this firsthand. A complex-procedure contact center had relied on tribal knowledge for years. 80% of calls and requests were escalated to supervisors, and new hires took up to a year to reach full proficiency, even after a 16-week training program.
After replacing memorization with guided, in-the-moment procedures:
- Escalations dropped from 80% to 10%, an 88% reduction
- New hires started contributing in 3 weeks instead of 16
- Full proficiency dropped from a year to 6 weeks
One leader described the shift simply: it felt like they didn't have new employees anymore.
How a Utilities Company Reduced Escalations from 80% to 10% With Guided Decision Trees
Learn how a California public utilities company replaced tribal knowledge with guided knowledge, cutting escalations from 80% to 10%, accelerating training, and boosting retention.
The Fix: Make the Implicit Explicit
A good decision tree takes what an expert holds in their head, the questions to ask, what the answers mean, what to do next, and makes it visible and followable for someone who isn't an expert yet.
That's the actual shift. You stop asking the employee to memorize the procedure. You teach them a different, much smaller skill: how to use a decision tree. And that decision tree lives in the moment of performance, on the call, with the customer in front of them, not as something they practiced once in a classroom and never touched again.
Three Doubts Leaders Raise, and What the Data Says
After working with organizations for over 2 decades, we've heard and seen it all. Here are three of the most common concerns leaders raise, and what our data has shown time and time again.
"Won't my people work faster if they're not using the decision tree?"
Data from multiple financial services and healthcare contact centers says the opposite. Employees following decision trees were twice as productive as employees working from memory. Memory-based work comes with backtracking and self-doubt. Decision tree work is deliberate, consistent, and confident.
"This will make my employees sound like robots."
A decision tree is sheet music for a musician. Following sheet music doesn't stop a musician from sounding exceptional. It guides the performance. It doesn't replace the performer.
"I don't want employees who just follow steps. I want them to actually understand what they're doing."
This is the concern that matters most, and it's also where the data is most counterintuitive. Across multiple implementations, employees trained to use decision trees show higher critical thinking than employees trained directly on procedures.
Here's why: decision trees free up the training time that used to go toward memorizing steps. That time goes to foundational knowledge instead, the what and the why behind the job. Employees start seeing how the different paths connect on their own, instead of just executing steps someone handed them.
In practice, that looks like an employee who can explain why a step exists instead of just following it, and who adapts to a policy change without needing to be retaught from scratch. Understanding grows faster, and performance improves right alongside it.
The Question Worth Asking
Every leader in a regulated industry should run their own procedures through this test: does this even meet the criteria to be trained into memory? Is it simple enough, frequent enough, stable enough?
If the answer is no, and for a meaningful share of your procedures, it will be, no amount of redesigning the training program fixes it. You don't need better training. You need a performance tool.
