Six weeks after a two-day programme in a Noida plant, a production manager told us the training “did not stick”. He was not being difficult. He had watched three supervisors go back to their old way of running the morning huddle within a fortnight. The programme had been well rated. The participants had enjoyed it. Both things were true at the same time.
That gap is not a facilitator problem and it is not a motivation problem. It is a design problem, and it is almost always decided before anyone walks into the room.
The workshop is the easy part
A two-day behavioural programme can create genuine intent. People leave with a plan, a phrase they liked, and a real resolution to handle the next difficult conversation differently. Nothing about that intent is fake. It simply has a short half life, because the environment they return to has not changed at all.
The participant goes back to the same queue, the same escalation route, the same manager who has never been briefed on what was taught, and the same twenty-seven open issues. The new behaviour is now competing with a system that was working fine before the programme existed. It loses.
Three things break transfer, and they break in this order.
No cue. Nothing in the workflow prompts the person to use the new behaviour at the exact moment it matters. They learned a structure for a difficult conversation, but the difficult conversation arrives at 4:40 pm on a Friday with four other things happening, and the structure is not in front of them.
No consequence. The old behaviour still gets the work done. The supervisor who simply tells people what to do gets the line running. The one who tries the coaching model spends three extra minutes per person and gets asked why the output dropped. The organisation rewards the thing it says it does not want.
No feedback. Nobody tells them whether they did it well. They try the new approach once, it feels awkward, nobody comments, and they quietly conclude it does not work. Adults do not persist with behaviour that produces no signal.
What actually holds
Programmes that survive six weeks have a follow-up structure agreed before the workshop, not after. It does not need to be elaborate. It needs to exist on a calendar with a name against it.
The version we use has five checkpoints. It is the same structure we write into every programme design. Each one is small enough that a busy manager will actually do it.
| When | What happens | Who owns it |
|---|---|---|
| Two weeks before | Manager identifies one real situation the participant will handle differently | Reporting manager |
| During the session | Participant commits to that specific situation, not a general intention | Participant |
| Week one | Participant runs the situation and writes three lines on what happened | Participant |
| Week three | Manager reviews those three lines in a normal one-to-one, fifteen minutes | Reporting manager |
| Week six | Short re-assessment or observed task, scored against the same anchors used in the session | L&D |
The critical row is the first one. If the manager cannot name a real situation before the programme, the programme has no target and the rest of the structure is decoration. We ask for it in the pre-work and we do not schedule the session until it arrives.
The second critical row is week three. It costs fifteen minutes of a one-to-one that was already happening. It is also the only point where someone with authority tells the participant whether the new behaviour was visible. Without it, the participant is guessing.
Transfer is a design decision made before the session, not a hope held after it.
A worked example
A mid-sized manufacturing client ran a two-day programme on structured problem solving for forty supervisors. The first cohort, run the conventional way, produced a satisfaction score of 4.5 out of 5 and no measurable change in the weekly defect review notes.
For the second cohort we changed four things and nothing else. The programme content, the facilitator and the duration stayed the same.
- Each supervisor arrived with one live defect from their own line, written on a card.
- The plant head sent a two-line note to all reporting managers explaining what they would be asked to review in week three.
- The week-three review used the same five questions taught in the session, so the manager did not need to learn a new framework.
- The defect review template was changed to require a root cause line in that same five-question format.
The measurement at week six is the same idea we describe in four levels of evaluation, applied to one programme instead of a portfolio.
At week six, 31 of the 40 supervisors had a completed root cause note in the system, against 6 of 40 in the first cohort. The fee was identical. The difference was roughly four hours of management time spread across six weeks.
These numbers illustrate the pattern we see. They are not a controlled study. Treat them as a reason to test the structure on one cohort, not as a benchmark.
The honest limit
Transfer design cannot fix a broken system. If the incentive structure rewards volume and punishes quality, no amount of week-three review will hold a behaviour that costs the participant money at appraisal time. If the person’s manager changes every quarter, the review will not happen.
In those cases the training is not the intervention. The incentive, the staffing or the reporting line is. Saying so is uncomfortable in a proposal, and it is cheaper than running a programme designed to fail.
There is also a real ceiling on how many behaviours one programme can move. One to three specific behaviours, practised in the participant’s own context, is a realistic target. A list of twelve competencies addressed in two days is a list of twelve things that will not happen.
The three-week wall
In practice the drop happens earlier than six weeks. Around week three, the participant hits the first genuinely difficult version of the situation they practised, does it badly, and has no one to debrief with. That is the moment the old behaviour comes back, and it is entirely predictable.
A transfer plan that survives this looks specific on paper. It says who reviews what, on which date, and what happens if the review does not occur. A plan that does not survive looks like this: “participants will be encouraged to apply the learning in their day-to-day work, supported by their reporting managers”. That sentence has no date, no name and no artefact. It is a wish with a full stop.
Three questions separate the two. What is the artefact the participant produces in week one? Who reads it? What does the reviewer do if it is missing?
What to do on Monday
Ask one question before approving any behavioural programme: what will be different in the participant’s calendar in week three?
If the answer is nothing, the programme is an event, not a change. If the answer is a fifteen-minute review with a named manager and a specific situation, you have a design with a chance. That question costs nothing and usually saves a quarter of the training budget.