Executive Essay · The Corporate Training Trap
An executive essay on why capability does not survive re-entry — and how to audit the conditions before you invest
Someone comes back from a program on delegation, on feedback, on working across functions, convinced. Within weeks, nothing is different, and the problem the program was bought for is still on the desk.
Most organizations can tell you what they spent on developing their people last year. Very few can tell you what happened to it.
They know how many attended. They know the programs were rated well. What they cannot say is whether anyone is doing, today, at work, the thing the program taught. When they look, they usually find that for a few weeks something changed and then the earlier way of working came back.
The usual reading is that the program was not good enough or the people did not follow through. Both readings point at the two things that were examined: the teaching and the participant. Neither looks at the one thing that was not examined, which is the place the participant went back to.
For capabilities that depend on judgment and on other people — leading, deciding, coordinating, speaking plainly — that place decides the outcome. What follows is a working model of how it decides, at four scales, with one check at each that can be run before a program is approved.
The individual: can, may, and has room
Start with the individual, because that is where the failure shows and where it is usually assumed to live.
An individual who does not use what they were taught is in one of three situations, and the three look identical from outside. They cannot: they did not learn it well enough. They may not: they learned it, and using it would cost them something where they work. Or they have no room: they learned it, nothing forbids it, and there is no hour in their week in which a slower, unpracticed way of working can be afforded.
Only the first is a matter of skill, and only the first is helped by more teaching. The second and third are conditions. Teach them again and you have spent the budget a second time on the one cause that was not in play.
The reason this is so often misread is that the individual tends to accept the first explanation. Capable people prefer a shortcoming they can work on to a situation they cannot govern. Ask them why the new practice lapsed and they will tell you they should have been more disciplined.
Audit implication: before attributing a lapse to the individual, establish which of the three it is. Ask what happened the first time they tried it — who responded, what it cost in time, what followed. If they cannot describe a first attempt, the issue may be skill. If they can, it is not.
The group: who must respond differently
Move up one scale. Almost nothing an individual is sent to learn can be done alone.
Delegating requires someone who accepts the decision and carries it. Listening requires someone who speaks. Disagreeing openly requires others who can be disagreed with and stay at the table. The trained individual was taught one half of a transaction. The people who hold the other half were taught nothing and told nothing.
So the group goes on expecting what it has always expected. It brings the usual questions to the usual person and wants the usual answer. No one opposes the new conduct. It simply does not fit the arrangements everyone else’s work is built on, and those arrangements exert a steady pull back toward the familiar. One position pulls harder than the rest: the person the individual answers to. One of the earliest field studies on the subject found that trained supervisors kept the new way of managing only where their own superiors managed that way. The rest went back.
Much of what is taught is also a risk. Saying that a plan will not work, or admitting an error early, is safe or unsafe depending on what the group did the last time someone tried. That history is not changed by training one of its members.
Audit implication: for each conduct you intend to develop, name who else has to behave differently for it to work — and start with the person they answer to. If those people are not part of the intervention, you are developing half of a transaction.
The organization: what is counted, who decides, how much is asked
At the next scale the expectations are written down. They are in what the organization measures, in who is permitted to decide, and in how much it asks of its people.
What is counted. People do what they are measured on, and they are right to. Someone measured on their own unit’s result and taught to cooperate across units will cooperate as far as that result can spare. When what is taught and what is counted disagree, what is counted wins, and the outcome is recorded as a failure of the teaching.
Who decides. A capability is the power to do something, and in an organization that power is assigned in advance. Someone taught to take decisions who returns to a role where the decision belongs two levels up has the skill and not the right. Holding them answerable for outcomes they do not control is not accountability. It is blame arranged beforehand.
How much is asked. This is the condition most often left out. Any new way of working is slower at first and needs to be done badly before it is done well. That requires margin: time and attention not already committed. In most large organizations the margin has been removed, deliberately, as waste. Near full load, a small extra cost does not cause a small delay; it causes the schedule to fail. So the new practice is dropped, rationally, by people who would have preferred to keep it. The organization asked for change and took away the only condition under which change happens.
There is a further effect. Under pressure, time is taken from improving and given to delivering. Capability thins, results slip, and the cause looks, to those in charge, like people who are not performing. Pressure is added, which thins capability further. The diagnosis keeps being confirmed by evidence that the conditions themselves produce.
Audit implication: set the conduct you want beside three documents — the performance measures, the decision rights, and the current workload of the people concerned. If any of the three contradicts it, change that first. And ask the capacity question in plain terms: what work will stop so that this can be practiced? If the answer is “nothing,” the program is being added to a full load and will not hold.
Society: what was let in
The widest scale is the hardest to see from inside, and the least measured. It is still at work.
Little of the pressure inside an organization starts there. The date behind the plan comes from a customer; the customer’s date from its own commitments; the target from what those who finance the business expect. What people assume about how to address those they answer to was formed long before they were hired, and differs from one country to the next more than internal programs usually allow for. A way of leading that is admired in one place can look weak or presumptuous in another.
Even the remedy often arrives from outside. Development programs spread through an industry because respected peers run them, not because each adopter established that it had the problem they address.
What matters for the executive is that the organization is not a pipe. It can pass outside pressure inward unchanged. It can add to it, by setting the internal date ahead of the external one. It can absorb part of it, in staffing and in time. Or it can change how it works so that the same outside requirement is met by an arrangement its people can sustain. Most organizations pass it on by default and add to it under stress, and describe both as necessity.
Audit implication: for the pressure that is defeating the conduct you want, separate what was truly imposed from outside from what was decided inside and described as imposed. The second category is usually larger than expected, and it is the part you can change.
The four at once
These are not four stages. They are present together in a single ordinary moment.
Someone tells the person they answer to, in the weekly meeting: this date cannot be met. They were taught to say such things. Whether it does any good depends on whether they have the judgment and the nerve; on what the others in the room do in the next two seconds; on whether anyone present is permitted to move the date, and whether any slack was left in the plan; and on the contract behind the date and what the surrounding culture makes of someone who contradicts the person they answer to in public.
Remove any one and it is a different event. Yet when the date is missed, the account of what happened will name one person. That is how a condition created at four scales comes to be treated as one individual’s shortcoming — and how the next program gets commissioned.
What is left in your hands
None of this argues against developing people. Where a capability is missing, it must be taught, and where the conditions support it, teaching works and lasts. The argument is about order.
For the individual: establish whether it is cannot, may not or no room before prescribing anything.
For the group: develop the people who must respond, beginning with the person the individual answers to, not only the individual who must act.
For the organization: align what is counted and who decides with what you want, and take work away before adding learning.
For society: distinguish what you were handed from what you chose, and decide how much of the first you will pass on.
Training comes after that. By then the measures, the decision rights and the workload no longer take the new conduct back, and what is taught has somewhere to stay. Put first, and alone, it is charged with every change that was not made, and it fails under a weight that was never its own.
The costs of the usual order do not vanish. They are passed down, scale by scale, to the person at the end of the line, who pays first. Then they come back up — in people who leave, in capability that thins behind good figures, in commitments that were never realistic. They return later and larger than the saving that caused them, and under another name.