How do you add an application step to an existing training program?

The most common objection to adding an application step is a scheduling and budget reflex: the program already exists, it took real money and political capital to build, and nobody wants to reopen it. But a program without an application step is an investment waiting to be lost: most of the knowledge it builds decays before it ever changes behavior. Retrofitting the application step is not adding a new project; it is completing the existing one, so the original investment produces the behavior change it was funded for. The program stays intact, the work is closer to an editorial project than a construction project, and this guide is the operational playbook for doing it.

What do you already have, and what is actually missing?

Start with an inventory of what exists, because it is more encouraging than most designers expect. The courses, e-learning modules, videos, slide decks, and job aids in an existing program are knowledge acquisition assets, and they generally work at what they were built for. Learners finish them knowing the material. The LMS distributes them, tracks them, and reports on them reliably. None of that needs to change.

What is missing is the step where learners learn to apply what those assets taught: a place to face situations that resemble real work, make consequential decisions, and get expert guidance at the moment of choice. Without that step, the knowledge decays before it changes behavior, a pattern documented in depth in this publication's guide to the training transfer gap. The retrofit does not fix the existing assets, because they are not broken. It completes them.

That single reframe, completion rather than replacement, drives every decision in the playbook that follows: which program to start with, where the new material goes, how it is measured, and how it is sold internally.

Step 1: Which program should get the application step first?

Almost any training is a candidate, because almost everything an organization teaches, a procedure, a protocol, a product, a sales process, a coaching model, is meant to be used somewhere, and anything meant to be used needs a place where people learn to use it. The application step does not require difficult or dramatic material: "where do you apply this?" is a scenario, and often a simple one. So the question is not which programs deserve an application step. It is where to start, and the answer is return on investment: which programs, left unapplied, cost the organization the most. Three signals identify that starting point:

1. The cost of non-application. Ask what it costs when graduates do not use the program on the job: compliance exposure, lost deals, rework, safety incidents. The programs where that number is largest are where the application step pays back fastest.
2. Visible complaints. Listen for the phrase "they know it, but they do not do it," in any of its local variants: "the training did not stick," "they revert under pressure," "they can recite the framework but freeze in the room." Every such complaint is a manager reporting the missing step in plain language, and a program that generates them arrives pre-sold.
3. Measurable stakes. Prefer a program whose outcomes someone already tracks: deal conversion, escalation rates, audit findings, guideline adherence. When the application step works, you want the improvement to land somewhere a stakeholder is already looking.

A program that scores on all three is the pilot. Lower-stakes programs are not excluded; they take a lighter-weight version, fewer scenarios and simpler decisions, and reap the same benefit. The first retrofit simply has to build the internal case for every retrofit after it.

Step 2: How do you identify the situations where people actually struggle?

The raw material of an application step is not the course content. It is the set of on-the-job situations where the course content fails to show up. Finding them is interview work, and the interviews are with the people closest to performance: managers, trainers, subject matter experts, and the people currently handling the situations well.

Three questions do most of the work:

• Where do trained people go wrong? Ask managers for the last three times a trained employee handled a situation badly. Specific incidents beat general impressions; "she discounted too early in the renewal call" is scenario material, "they need better negotiation skills" is not.
• What do the people who handle this well do differently? Experts often cannot recite their own reasoning in the abstract, but they can walk through a concrete situation and narrate the choice: what they noticed, what they ruled out, why the tempting option fails. That narration is the expert guidance the scenarios will deliver. Anyone who can articulate the reasoning behind the strong choice can be a source: a trainer, an experienced practitioner, a subject matter expert.
• Which situations recur? A scenario is worth building when the situation it trains arrives regularly. Cathy Moore's action mapping method makes the same discipline explicit: start from the business outcome, list the on-the-job behaviors that produce it, and design activities only for those behaviors (Moore, action mapping).

The output of this step is a shortlist, typically a handful of situations, each with a pivotal decision, several plausible ways to get it wrong, real consequences, and a named person who can explain the strong choice. If a candidate situation lacks any of those four properties, cut it.

Step 3: How do you convert situations into scenario experiences?

Each shortlisted situation becomes a scenario experience, and three design properties separate a real scenario experience from a repackaged quiz.

A realistic situation. The scenario has to feel like the job, not a simplified version of it: the same messy inputs, conflicting priorities, and options that sound reasonable until a person thinks them through. This is not cosmetic. Wilson Learning's synthesis of 32 transfer studies, a vendor review of the research literature, found that the closer in-course experiences replicate real working conditions, what the literature calls fidelity, the greater the transfer to job performance (Leimbach, Learning Transfer Model). A scenario assembled from clean, pre-simplified inputs only proves learners can handle clean, pre-simplified inputs, which is not the ability the job actually requires.

Consequential decisions. The learner chooses among options that all look defensible, and what happens next follows from that choice rather than from a rubric. The stakeholder pushes back, the patient's condition evolves, the negotiation shifts. What makes a decision consequential is not a dramatic stake but an outcome the learner has to live with: they find out where the choice leads instead of being told it was wrong.

Expert guidance at the moment of choice. When the learner picks a weaker option, the expert reasoning gathered in Step 2 shows up immediately, attached to the choice the learner just made: why the option was tempting, why it falls short, and what the stronger move looks like. Research on how expertise develops consistently points to immediate, informative feedback on each attempt as the condition under which judgment improves (Ericsson, 2008); a score delivered after the module has moved on has nothing left to correct.

Keep each scenario tight: one situation, a few pivotal decisions, guidance at each. Depth per decision beats breadth per scenario, and a short scenario that forces real judgment outperforms a long one that tours the content.

Step 4: Where does the application step go in the learner's journey?

Placement follows one rule: the scenario comes after the knowledge it draws on, close enough that the knowledge is still fresh. Within that rule, three patterns cover most programs.

Interleaved. For multi-module programs, place a scenario after each major knowledge block. Learners alternate between acquiring and applying, which keeps early modules from decaying while later ones are still being taught. This is usually the strongest pattern, because it gives learners repeated cycles of deciding and being guided.

Capstone. When interleaving is impractical, a scenario sequence at the end of the program still completes the journey, and it doubles as a summative checkpoint: the last thing the program observes is the learner deciding, not the learner recalling.

Refresher. The same scenarios earn a second deployment weeks or months later. The decay documented by Saks and Belcourt is steepest in the months after training (Saks & Belcourt, 2006), and a reassigned scenario meets that window with application rather than a content re-read; on platforms built for reuse, such as AliveSim, redeploying an existing scenario adds little to no new build work.

In every pattern the existing assets keep their place in the sequence. The retrofit adds items to the learner's path; it removes nothing.

Step 5: How do you measure the application step from day one?

Measurement is where retrofits most often shortchange themselves, because the temptation is to launch first and instrument later. Later never has a baseline.

The essential discipline is to capture learners' first choices in each scenario before any guidance is delivered. Those baseline decisions are the closest thing a program can get to observing what learners would have done on the job, and they are the comparison point for everything that follows. From there, two changes tell the story:

• Change within the experience. After guidance, does the learner's decision quality improve on subsequent choices and scenarios? First-choice quality improving from scenario to scenario means learners are carrying reasoning forward, not being corrected anew each time.
• Change across the cohort. Where did the group's baseline choices cluster? Which decisions improved most after guidance, and which stayed stubborn? Stubborn decisions are a signal to the program: either the scenario needs sharper guidance or the knowledge modules upstream need reinforcement.

This is also where tooling matters, because decision-level capture is tedious to improvise. Whatever the tool, the principle stands: if the application step cannot show what learners chose and how those choices changed, it will be defended with the same completion metrics that failed to defend the original program.

Purpose-built platforms make that discipline easier to sustain. AliveSim's Guided Scenarios, for example, wrap a program's existing knowledge assets in realistic decision situations with expert guidance at the moment of choice, capturing each learner's baseline and subsequent choices automatically rather than replacing anything underneath. That is the retrofit principle this playbook describes, built into the tool instead of left to a facilitator's memory. AliveSim

How do you handle the organizational realities?

A retrofit succeeds or stalls on three conversations that have little to do with learning design.

Positioning to stakeholders. Lead with completion, not novelty. The organization already invested in the program; the application step is what makes that investment pay. This framing matters because it changes what the proposal competes with: an addition that completes existing spend is weighed against the decay of that spend, while a "new learning initiative" is weighed against every other new initiative. The method itself needs little defense; as of ATD's 2021 research, 98% of surveyed organizations already used scenario-based learning in some form (ATD, 2021). The case to make is not that scenarios work, but that this program is incomplete without them.

Budget framing. Budget the step as an addition to the existing program line, sized against the value of the investment it protects, not as a standalone project competing for new funds. A modest addition that rescues a large sunk cost is one of the easier cases in a training budget cycle.

Keeping the LMS workflow intact. Nothing about the retrofit should disturb how the program is administered. Scenario experiences link from the LMS like any other assignment, completion flows back into existing tracking, and on browser-based platforms such as AliveSim there is little new infrastructure for IT to evaluate. Administrators keep their reporting; the program simply gains a new kind of evidence inside it. The less the retrofit changes operationally, the faster it clears review.

What are the common mistakes?

Three failure patterns account for most weak retrofits.

Bolting on a quiz and calling it application. Rewording a multiple-choice question in scenario language, a customer's name, a line of dialogue, does not change which ability gets exercised. Watch what happens after the learner clicks: if nothing unfolds and the program simply marks the answer right or wrong, what is being measured is whether the learner can pick the correct option off a short list, and that is still recall, not application.

Scenarios about trivia instead of judgment. A scenario that asks the learner to recall a product specification or a policy clause inside a story is still testing recall. The decision at the center of each scenario should be one where trained people genuinely go wrong for reasons of judgment: they misread the situation, choose the tempting shortcut, sequence the conversation badly. If every learner would get it right on the first attempt, the scenario is teaching nothing that the modules did not.

Skipping baseline measurement. Launching without capturing unguided first choices forfeits the strongest evidence the retrofit will ever have. The baseline is only observable once per learner; there is no going back for it. Instrument before the first cohort, not after.

Avoid those three, follow the five steps, and the retrofit stays what it should be: a focused addition that completes a program the organization already believes in, with evidence that the completion worked.

References

• ATD Research. (2021). Use of simulations and scenario-based learning is rising. Association for Talent Development.
• Ericsson, K. A. (2008). Deliberate practice and acquisition of expert performance: A general overview. Academic Emergency Medicine, 15(11), 988–994.
• Leimbach, M. Learning transfer model: A research-driven approach to enhancing learning effectiveness. Wilson Learning Worldwide.
• Moore, C. Action mapping: A visual approach to training design. cathy-moore.com.
• Saks, A. M., & Belcourt, M. (2006). An investigation of training activities and transfer of training in organizations. Human Resource Management, 45(4), 629–648.

Published April 22, 2025 · Updated July 15, 2026 · 12 min read