Field Series · General · Report
The Corporate Training Trap
Why the problems you trained for are still on your desk.
The program ran. People rated it highly. And the problems you bought it for are still there.
Bad news still arrives late. Functions still do not cooperate. Decisions are still made twice. The next proposal in front of you is another program.
The organization responds to a problem in the architecture of work by trying to repair the people working inside it.
Training is necessary. Alone, it is insufficient.
The finding
Some skills travel. Others do not travel alone.
Skills that travel
Operating a system, following a procedure: teach it, and it is used.
Skills that need a response
Giving honest feedback. Delegating a decision. Disagreeing with whoever holds the decision. Cooperating across a boundary. Stopping a project that is failing. Each one needs somebody else to respond differently.
The capability is formed in the person. The occasion to use it is granted by others.
Key figures
What the evidence shows
What people return to decides whether training is used.
For leading, deciding and communicating, the workplace weighs about six times more on whether training is applied than it does for software and procedures, where it barely matters.
→ You can train the person perfectly and still get nothing back.
The person you answer to counts twice as much as your colleagues.
Whether people use what they learned depends more on the person they answer to than on anyone else around them.
→ A program that trains the group and leaves out the person they answer to funds the weaker half.
The higher the participants sit, the less the training changes.
Leadership training changes what people do at work about five times more at the lower levels of an organization than at the top.
→ Your most expensive programs move behavior the least.
Redesign first, train last.
One division that changed roles and accountability before training launched nine new products in two years, against five in the previous four.
→ The order of the steps decides the return.
Five lenses
Why the trap holds
Open capabilities are not used alone
A program delivers half of a transaction. The other half — the person they answer to, the colleagues, the role — was taught nothing.
The diagnosis goes to the wrong address
A problem the system produces is read as a skill someone lacks. Training is the remedy that is easiest to commission: it can be bought, scheduled and counted, and it leaves everything else as it is.
The same pattern at four scales
Individual. Group. Organization. Society. The symptom shows in the person, but each of the four contributes to the problem, in its own way and to its own degree. Fix one and leave the other three, and the three undo it.
No room
A new way of working is slower at first. In a week that is already full, a small added cost does not cause a small delay. It breaks the week. Dropping the new way is the rational choice. Margin, not heroics.
The system takes it back
The earlier conduct returns because the arrangement that selected it never changed. It is not a relapse. It is the design, working.
No single culprit
Not the person or the system. All four, each in its measure.
This is not a choice between blaming people and blaming structures. The individual, the group, the organization and the society around it are all part of the problem — each in its own way and to its own degree. That is why each is also part of the solution.
The person still has to learn and to act. The group has to respond. The organization has to make room and pay for the conduct it asks for. And the pressure that comes from outside has to be met, not simply passed on. Work on one of the four alone, and the other three take it back.
Run the diagnosis
The usual reading, and the question behind it
Pick what you see. Read the usual diagnosis first — then reveal the second reading.
The purest case
One sentence, four scales
Someone tells the person they answer to, in the weekly meeting:
“This date cannot be met.”
That person was trained to say such things. Whether it does any good depends on four things at once.
Whether the person has the judgment and the nerve.
What the others in the room do in the next two seconds.
Whether anyone present may move the date, and whether any slack was left in the plan.
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. When the date is missed, the account will name one person. All four were part of it — and all four are part of the fix.
What changes
The other order
| The usual order | The other order | |
|---|---|---|
| First question | What do our people lack? | What would meet them if they did it? |
| Who is addressed | The individual, away from the work | The group and its chain of command, in the work |
| Capacity | Assumed | Released first |
| Training | First and only step | Last step |
| Proof | Attendance and satisfaction | What people do at work |
Diagnose first. Redesign what blocks it. Make room. Train last.
Why not simply buy better training?
The problem is on both sides of the door
On this side, the teaching itself is often built for the wrong kind of skill. Capabilities that depend on judgment and on other people are taught as if they were procedures: a model, a script, a role-play, a certificate. The method is wrong, the instruments measure the wrong thing, and the timing is wrong — away from the work, in one block, long before or long after the moment the skill is needed.
On the other side is everything the teaching has no authority over: the person they answer to, the measures, the decision rights, the load.
A better program of the same kind fixes neither. What works is a different kind of training — in the work, over time, with the people who must respond — and a different mandate for what lies beyond it.
Architect’s Essay
A Better Key
Why trained people still cannot use what they were taught. A key cut from a template, for a door barred from the inside.
The full report
The Corporate Training Trap
The complete argument, the evidence behind every figure on this page, and the full list of references.
By field
See how it applies
More fields to come. Programs: diagnosis, intervention, transfer →
Sources for the figures on this page.
Beer, M., Finnström, M. & Schrader, D. (2016). The great training robbery (Harvard Business School Working Paper No. 16-121).
Blume, B. D., Ford, J. K., Baldwin, T. T. & Huang, J. L. (2010). Transfer of training: A meta-analytic review. Journal of Management, 36(4), 1065–1105.
Lacerenza, C. N., Reyes, D. L., Marlow, S. L., Joseph, D. L. & Salas, E. (2017). Leadership training design, delivery, and implementation: A meta-analysis. Journal of Applied Psychology, 102(12), 1686–1718.