Your company spends real money on leadership development, and you know the pattern. The cohort comes back from the program energized, armed with new vocabulary, genuinely intending to lead differently. Six weeks later, nothing has changed.
The reflex is to blame the program and shop for a better one. Before you do, watch what the returning manager actually walks back into.
They were taught be decisive in a company that requires six approvals. They were taught empower your team in a shop where the boss still holds every call. They were taught move faster in a culture that wants everyone briefed, pre-briefed and emotionally prepared for the meeting before the meeting.
The program delivered fifty hours of instruction. The environment delivers two thousand hours of the opposite lesson, every year, with better attendance.
The environment is the real curriculum
By environment I mean your operating system: decision rights, approval counts, what gets rewarded, what gets punished, and what the most senior person in the room actually does on a Tuesday. That system teaches every single day, and it never misses a session.
Repetition out-teaches instruction, and pressure settles the argument. One write-up I read cites a 2009 study by Schwabe and Wolf in the Journal of Neuroscience reporting that stress reliably pushes behavior toward the automatic and the habitual, and away from the deliberate response somebody just learned. Under pressure, your newly trained leader does not run the new material. They run the groove your company has been carving for years.
Let me be careful about what I am claiming. I am not telling you most programs fail. I don't have a number for that, and I won't invent one. The failure worth governing is narrower and more specific: the program taught a behavior, and the environment rewards the opposite. When those two collide, bet on volume.
This is not a complaint that only skeptics make, either. One program provider's own diagnosis concedes it: development often ends the day the teaching does, leaders leave inspired and then struggle to name one thing that has actually changed in how they lead, and once the final session closes, turning ideas into behavior is left to the individual. Vendors will happily sell you follow-up support for that gap. What none of them can sell you is the room your manager walks back into. The new behavior runs through your operating system, or it doesn't run.
The training-transfer research has been making the same point for a while. One write-up of that literature I read summarizes a 2010 meta-analysis by Blume and colleagues, eighty-nine studies in the Journal of Management, and reports the headline: a supportive work environment is among the strongest predictors of whether trained skills ever show up on the job. The same write-up cites a 2020 study by Hughes and colleagues, in Human Factors, finding peer and supervisor support strongest for sustainment, meaning whether the skill is still in use months later, and it draws the conclusion plainly: the people around the learner determine whether the skill survives. Your manager's new behavior lives or dies in the place they work, not the place they were trained.
Two qualifications, before this becomes a slogan
First: some programs are simply bad. The environment cannot become the universal excuse that protects a weak purchase from evaluation. If the curriculum was vague, the facilitator was reading slides and the case studies came from somebody else's industry, that is a program problem, and no amount of environmental tuning rescues it.
Second: some skills genuinely have to be taught before any environment can hold them. Giving hard feedback, reading a P&L, staying calm in an escalation: nobody absorbs those by osmosis. Programs are not useless. The claim is narrower, and it is the whole point: a program without an environmental change to hold what it teaches is not development. It's an event. When development does work, it looks like this: real decisions with real stakes, built into the daily work, measured by how the team performs rather than by who finished the module.
The buyer owns the environment
Here is the uncomfortable part. The environment is not weather. Somebody owns it. The approval count is a policy. The metric that quietly rewards the old behavior is a choice. The meeting before the meeting is a norm. And what you do on a Tuesday is you.
That is why shopping for a better program is usually the wrong first move: it buys the solution from somebody who does not own the problem. What you cannot buy is the environment the new behavior has to survive in. No vendor owns your approval count, your metrics or your habits; you do. Which also means the audit below points at you before it points at anyone else.
The Return-Trip Audit
Run this on your most recent development investment before you approve the next one. It's a proposed practitioner exercise, not a studied protocol, and it takes one sitting plus two small follow-ups.
Step zero comes first, because it protects the audit from its biggest failure mode. If the program cannot name the two or three specific behaviors it was supposed to teach, in plain verbs, stop. Evaluate the program before you audit anything else. An audit needs a target, and a vague curriculum is a vendor problem, not an environment problem.
First, write the behaviors down. Plain verbs, not values: decides the pricing exception without routing it to my inbox; gives the underperformer the real review instead of the softened one; makes the call in the meeting instead of scheduling the meeting before the meeting. If the program taught empowerment as a noun, translate it into something you could watch a person do.
Second, for each behavior, name what in your operating environment currently rewards the opposite. Not culture. Culture is how this audit dies. An approval count. A metric. A meeting norm. Your own habit. Be specific: decisiveness died because decisions above a rounding error need six signatures, three of which sit in your inbox.
Third, pick the one contradiction with your name on it and change it within thirty days of the cohort's return. Thirty days, not the next planning cycle. A cohort that watches its sponsor keep the old behavior learns faster than any classroom taught it.
Fourth, at week two, ask one returning participant: what did the environment teach you this week? Treat the answer as data, not disloyalty. Your people are the audit's sensors, and punishing the message turns the sensors off.
One line for the next proposal, deliberately the smallest part of this exercise: when it arrives, it names, on the same page as the content, the environmental change that will hold the new behavior, or the money waits. That is a purchase decision about development, and it sits at the end of this list on purpose. The audit is the tool. The gate is just its last line.
If the audit comes back clean, if your environment contradicts nothing the program teaches, buy the next one with confidence. A clean result is the point of the audit, not a loophole. It means the next program has somewhere to land, and the landing is yours to provide.
Development budgets are being set right now, including the fresh wave of AI-fluency training being priced this quarter. If this year's cohort is already back, run the audit this week. The program ended. The environment is still teaching.
Working through this with your own team?