Devika Sen runs L&D for a speciality chemicals company in Kolkata. Last October she signed off on a two-day communication programme for 40 plant supervisors. The trainer was good. The feedback form came back at 4.6 out of 5. She forwarded the scores to her CHRO and closed the file.
In March, the same CHRO asked her a simple question. Had escalations from the shop floor come down? Devika did not know. Nobody had looked at a supervisor’s behaviour since the workshop ended.
That gap is a calendar problem. The workshop was the only thing anyone had scheduled. What decides whether a programme changes anything is the twelve weeks that follow it. In most Indian companies, nothing sits in that window at all.
A training reinforcement plan is a written schedule of what happens in the 90 days after a workshop, naming who does what on which day, so that learning turns into observed behaviour at work rather than a good feedback score.
Key Takeaways
|
Why training fades, and how fast
The strongest evidence on this comes from a study of 150 organisations by Alan Saks and Monica Belcourt, published in Human Resource Management. Training professionals reported that 62% of employees applied what they learned immediately after a programme. At six months that figure had fallen to 44%. At one year it stood at 34%.
Read that as a spend figure and it stings. Two-thirds of what a company pays for is gone within a year.
The reason is rarely the content. The workplace a person returns to has not changed at all. Their manager runs the same review meeting and rewards the same behaviour as the week before. A meta-analytic review of 89 transfer studies in the Journal of Management put a supportive work environment alongside ability and motivation as a consistent predictor of transfer. Support here means the manager, the review rhythm, and permission to do the new thing badly for a few weeks.
We have written before about why most leadership training fails to change behaviour at work. This post is the fix, written as a schedule you can run from Monday.
The Five Checkpoints
Day 0: The Contract
Fix this before anyone books a room. Agree three things in writing: the one or two behaviours the programme should change, who will observe them, and what the business should see if it works.
For Devika’s supervisors, the behaviour was specific. Raise a production issue to the shift in-charge within 30 minutes, in writing, with a proposed fix. Not “communicate better”.
Then ask the sponsor one question. What will look different in this department in three months? If nobody can answer, the programme is entertainment.
Day 7: The Manager Debrief
Within a week of the workshop, each participant sits with their manager for 20 minutes. Not HR. The manager.
Four questions do the work:
- What is the one thing you are going to do differently?
- Where will you try it first, and by when?
- What will get in your way here?
- What do you need from me?
That last question matters most. It moves the manager from spectator to sponsor. If your managers cannot hold this conversation, that is your real gap, and it is a separate piece of work. Our feedback training programme exists largely because of debriefs that go badly.
Day 30: The Application Check
At day 30 you look at one behaviour, observed, not self-reported. A self-rating survey here tells you how people feel about the training, not what they did.
Pick the cheapest available evidence. A manager sits in on two calls. Three escalation emails get read. In Devika’s case, her team pulled 20 shift handover notes and counted how many carried a proposed fix. The answer was six.
Six out of twenty is a useful number. It is honest, it is early, and it can still be moved.
Day 60: The Consistency Check
After first check that tells you whether people tried.
This next check tells you whether it survived a bad week.
This is where most reinforcement quietly dies. The novelty has gone and the quarter has got busy. Look at the pattern of slippage rather than the average. Usually two or three people have kept it up, most have drifted, and one or two never started.
Treat those three groups differently. The ones who kept it up become your internal examples, so say their names in a team meeting. The drifters need one nudge from their manager, not a refresher workshop. The ones who never started usually have a blocker nobody has asked about. It is far more often a process problem than a skill problem. We see this constantly in performance improvement work, where execution fails long before capability does.
Day 90: The Sustain Review
At 90 days you answer two questions and write down both answers.
First, did the behaviour shift? Compare it with the same evidence you gathered at day 30, measured the same way. Devika’s team pulled another 20 handover notes. Fourteen carried a proposed fix.
Second, what keeps this going without you? Reinforcement ends. Systems do not. Move the behaviour into something permanent: the appraisal form, the induction checklist, the weekly review agenda, the shift handover template. A behaviour that has not been written into a system by day 90 will not be there at month nine.
Who owns what
| Checkpoint | L&D / HR | The manager | The participant | The training partner |
| Day 0 Contract | Runs the conversation, records the agreement | Names the behaviour to change | Not involved | Designs backwards from the behaviour |
| Day 7 Debrief | Sends the four questions, tracks completion | Holds the 20-minute conversation | Commits to one action, one date | Supplies the debrief script |
| Day 30 Application Check | Collects the evidence | Observes and reports what was seen | Applies it, flags blockers | Reviews the pattern with L&D |
| Day 60 Consistency Check | Segments who kept it, drifted, never started | Nudges the drifters, unblocks the rest | Keeps going, asks for help | Runs a 60-minute follow-up (Capsule module) if needed |
| Day 90 Sustain Review | Compares day 30 and day 90 evidence | Signs off on what changed | Reflects, mentors a peer | Presents findings, recommends what sustains |
Learn → Apply → Measure → Sustain
A typical training company in India delivers a workshop, collects a smile sheet, invoices, and leaves. That is the industry norm, and it is why HR heads have stopped believing training numbers.
Excellential works to a named methodology instead: Learn → Apply → Measure → Sustain. It is built on the Kirkpatrick Model, so it reads familiar to any HR head who knows Levels 1 to 4. What it adds is the part Kirkpatrick describes and most vendors skip.
- Learn is the workshop. Day zero, not the deliverable.
- Apply is Day 7 and Day 30. Kirkpatrick Level 3 territory, made into a calendar entry with a named owner.
- Measure is the pre and post behavioural comparison at Day 90, using the same evidence both times.
- Sustain is what you write into the appraisal, the induction pack or the review agenda so it outlives the programme.
We have made the case elsewhere for why training scores mislead and what to measure instead. The 90-day plan is how that measurement gets collected. Our leadership programme for a life insurance client shows the full pattern, with sustain work running longer than classroom work.
Three clauses to put in your next training contract
Buying reinforcement costs less than retrofitting it. Ask for these three things before you sign.
- A day 30 and day 90 measurement, defined in the proposal. The method, the sample size and who collects it. Vague wording about “post-programme support” means nothing.
- A manager debrief pack. Scripts, the four questions, and a 45-minute briefing for managers before the workshop, not after.
- A written sustain recommendation at day 90. What goes into your appraisal form or review agenda, in the vendor’s own words, on paper.
If a vendor resists all three, you have learned something useful for the price of a conversation. Our post on the questions to ask before signing a training proposal goes further on this.
Indian HR leaders are already moving this way. Bhavna Batra, vice president of People at S&P Global, told People Matters in July 2026 that she reverses the usual order of learning evaluation. “For me, the first stage of impact in the world we operate in is business results,” she said. Decide what should improve before you design the programme. The 90 days afterwards then write themselves.
A workshop is a purchase. The ninety days after it are the investment.
Devika now runs every programme this way. It costs her four calendar entries and roughly six hours per cohort. If you are planning a programme and want the same structure around it, book a consultation and we will map the ninety days with you first.
FAQ
How long should training reinforcement last?
Ninety days, with checkpoints at day 7, day 30, day 60 and day 90. That window is long enough for a behaviour to survive a busy quarter, and short enough that people still remember the workshop. After ninety days, reinforcement should stop being a programme and become part of a system, such as the appraisal form or the weekly review agenda.
Do we need a learning platform or software to run a reinforcement plan?
No. A reinforcement plan runs on four calendar invites, a manager who holds a 20-minute conversation, and a spreadsheet with one observed behaviour per participant. Software starts to help above roughly 300 learners a year. Below that, a platform buys very little that a disciplined calendar does not already give you.
How do we measure behaviour change without a survey tool?
Use evidence that already exists in the business. Read a sample of emails, sit in on two calls, check handover records or count exceptions in a report. Take the same sample at day 30 and again at day 90, measured the same way. Observed evidence beats self-rated surveys, because people rate their own application generously.
What if we have no L&D team at all?
Then the sponsor of the training runs the plan, and the participants’ managers do the checkpoints. A founder can run a 90-day reinforcement plan for a 40-person company in under five hours. The one part you cannot skip is the day 7 manager debrief. That is where the workshop either enters the workflow or leaves the building.
Who is responsible for making training stick, the trainer or the manager?
The manager. A trainer controls two days. A manager controls the following ninety, sets what gets asked about in reviews, and decides what gets rewarded. L&D schedules the reinforcement and collects the evidence. A good training partner supplies the scripts and the measurement design. The behaviour itself lives or dies with the line manager.
What is the difference between reinforcement and a refresher session?
A refresher repeats content. Reinforcement changes the conditions the person works in, through manager conversations, observation, feedback and a change to a process or form. Refreshers help when knowledge has been forgotten. Most application failures are not knowledge failures, so a refresher often treats the wrong problem.
