You’ve got a budget review in two weeks. Your training data shows a 92% completion rate and a 4.2 out of 5 satisfaction score. You already know that’s not going to be enough. The CFO wants a dollar figure, and “people really liked the workshop” is not a dollar figure.
This is where most L&D leaders get stuck. Every training ROI guide assumes you have an analytics team, a fully integrated HRIS, and six months of clean data. Most mid-market companies have none of those things. So they default to reporting completion rates (which nobody trusts) or skip measurement entirely.
This guide gives you a tiered approach. If you can only do one thing, do Tier 1. If you have more resources, move to Tier 2 or 3. The goal is not a perfect calculation. It’s credible evidence that your training investment is working.
Why Training ROI Is Hard to Measure (And Why Most Guides Ignore This)
Before we get into frameworks, let’s be honest about why this is difficult. It’s not because L&D teams are lazy. Three structural problems make training ROI genuinely hard.
The attribution problem. A manager completes a coaching program in March. Their team’s retention improves in June. Was it the coaching? The new compensation policy? The fact that a difficult team member left in April? You can rarely isolate training as the single cause of a business outcome.
The lag problem. Training impacts behavior over weeks and months, not days. A leadership development program that starts in January might not show up in engagement scores until Q3. Budget cycles don’t wait that long. You need leading indicators while you wait for lagging ones.
The data access problem. Measuring business results (Kirkpatrick Level 4, which we’ll get to) requires retention data, performance ratings, and customer satisfaction scores. In most companies, that data lives in systems the L&D team can’t access or doesn’t exist in a usable format.
The honest reframe: your goal is not a perfect ROI calculation. It’s directional evidence strong enough to earn continued investment. CFOs don’t expect laboratory precision from HR. They expect a coherent story backed by reasonable numbers.
The Kirkpatrick Model: What It Is and Why Level 4 Is Out of Reach
If you’ve worked in L&D for more than a year, you’ve seen the Kirkpatrick model. Four levels of training evaluation, plus a fifth added later by Jack Phillips. Let’s walk through them honestly.
Level 1: Reaction. Did participants like the training? This is your post-training survey. Easy to collect, nearly universal, and almost useless for proving business value. A 4.5 out of 5 satisfaction score tells you the trainer was engaging. It tells you nothing about whether anyone will do anything differently on Monday.
Level 2: Learning. Did participants learn something? Pre-test and post-test scores, knowledge checks, skill assessments. Better than Level 1, but still doesn’t answer the question executives care about: did the learning change behavior at work?
Level 3: Behavior. Did participants change what they do on the job? This is where training ROI actually lives. Behavior change bridges “we taught them something” and “it improved business outcomes.” Most measurement efforts stop here, because tracking behavior requires observation, 360-degree feedback, or a system that measures skills over time.
Level 4: Results. Did the training improve business metrics? Revenue, retention, productivity, customer satisfaction. This is what your CFO wants. It’s also what 90% of L&D teams cannot measure directly, because isolating training’s contribution from every other variable is close to impossible without controlled experiments.
Level 5 (Phillips ROI). Convert Level 4 results into a monetary value, subtract program cost, express as a percentage. Conceptually simple. Practically, it depends entirely on how credible your Level 4 data is.
What You Can Realistically Measure at Each Company Size
| Company Size | Realistic Level | What You’ll Use | Typical Gap |
|---|---|---|---|
| Under 500 employees | Level 1-2 | Post-training surveys, quizzes | No behavior tracking system |
| 500 to 2,000 | Level 2-3 | Surveys plus manager observation or 360s | Limited HRIS integration |
| 2,000 to 10,000 | Level 3-4 | 360 feedback, skill platforms, some HRIS data | Attribution still fuzzy |
| 10,000+ | Level 3-4 (rarely 5) | Full analytics stack, control groups possible | Political barriers to data access |
The uncomfortable truth: most companies operate at Level 1. The ones measuring training ROI well are doing Level 3 with some directional Level 4 evidence.
A Tiered Measurement Framework That Actually Works
Forget trying to jump straight to Level 4. Here’s a three-tier approach that starts with what any company can do and builds from there.
Tier 1: Minimum Viable ROI (Any Company, Any Budget)
This takes about an hour to set up and works even if you have zero analytics infrastructure. Three steps.
Step 1: Define the business problem before the program starts. Not “we need leadership development.” Something specific: “First-year manager turnover is 34%, and exit interviews cite lack of support.” Write it down. Get your stakeholder to agree this is the problem you’re solving.
Step 2: Baseline the metric. Find the number today. Manager turnover: 34%. Time-to-productivity for new managers: 6 months. Team engagement score: 3.1 out of 5. You need a “before” number to have an “after” number.
Step 3: Measure the delta at 30, 60, and 90 days. Check the same metric at regular intervals. Did it move? You won’t be able to prove the training caused the change. But if you defined the problem clearly and the metric improved, you have a reasonable case. Directional is infinitely better than “92% completion rate.”
Tier 2: Behavior Change as Proxy (With 360 or Skill Tracking)
Level 3, behavior change, is the strongest leading indicator of business results. If managers are actually doing things differently after training, the business outcomes tend to follow. The challenge has always been measuring behavior at scale.
Traditional approaches use 360-degree feedback or manager observation surveys. Both work, but they’re slow and typically happen once or twice a year. Not enough data points to connect to a specific training program.
At Risely, we track behavior change across 83 skills continuously. Across 4,000+ users, the average improvement is 26% in targeted skills within 12 weeks. That number comes from repeated assessments, team feedback, and coaching engagement data, not a single post-program survey.
If you can measure behavior change at the skill level, you have something much more convincing than completion rates. A CFO might dismiss “94% of managers completed the program.” They won’t dismiss “managers showed a 26% improvement in delegation and feedback skills, as rated by their direct reports.”
Tier 3: Full ROI Calculation (High-Stakes Programs)
For expensive programs where you need to justify significant spend, you can attempt a Phillips-style ROI calculation. Walk through this worked example.
Scenario: You enrolled 20 managers in a 6-month coaching program. The program cost $30,000 total (platform fees, facilitator time, manager hours).
Outcome: In the 12 months after the program, teams led by coached managers had 3 fewer voluntary departures than the comparison group of uncoached managers.
Calculation:
- Replacement cost per employee: $60,000 (conservative estimate, typically 50-200% of salary)
- Value of reduced turnover: 3 departures x $60,000 = $180,000
- Program cost: $30,000
- Net benefit: $180,000 minus $30,000 = $150,000
- Training ROI: ($150,000 / $30,000) x 100 = 500%
That’s a clean number. CFOs like clean numbers.
But let’s be honest about the assumptions. We assumed the 3-departure difference was caused by coaching. Maybe it was. Maybe those teams also had better projects. The isolation problem never fully goes away.
Best practice: acknowledge the uncertainty. Say “even if coaching accounted for only half of the improvement, the ROI is still 200%.” Conservative estimates are more credible than precise ones.
The Metrics That Matter vs. The Metrics That Are Easy
Most L&D reports are full of easy metrics because hard metrics require more work. The difference matters.
| Easy but Weak | Hard but Strong |
|---|---|
| Completion rate | Behavior change at 90 days |
| Satisfaction score (Level 1) | Skill improvement over time (Level 3) |
| Hours of training delivered | Retention delta for trained vs. untrained groups |
| Number of participants | Manager effectiveness ratings from direct reports |
| Courses launched | Time-to-productivity for newly promoted managers |
CFOs have seen enough completion rate reports to distrust them instinctively. A 95% completion rate on mandatory training means people clicked through the slides. Everyone in the room knows it.
One strong metric beats five weak ones. If you can show that managers who went through your program received higher effectiveness ratings from their teams six months later, that’s a story a CFO can act on. Not a perfect ROI calculation, but evidence of value, which is what the conversation actually needs.
How to Build Your Training ROI Case in 30 Minutes
You don’t need a month-long data project. You need 30 minutes and a clear structure.
-
Write down the business problem in one sentence. Not “improve leadership skills.” Something like “reduce first-year manager attrition from 34% to under 25%.”
-
Identify the baseline metric. Find today’s number. If you don’t have it, that’s your first action item.
-
Define success before the program starts. What number, at what time horizon, would make this investment worth it? Get your executive sponsor to agree in writing. This is the single most important step.
-
Pick your measurement tier. Tier 1 for small teams with no analytics support. Tier 2 if you can track behavior change. Tier 3 if the program warrants a full ROI model.
-
Schedule the measurement. Put 30-day, 60-day, and 90-day check-ins on your calendar right now. Not “we’ll measure it later.” A specific date with a specific metric to check.
If you want help building a shareable business case with these numbers, Risely’s business case builder generates one tailored to your company size and industry. It takes about five minutes.
Training ROI Calculator: The Simple Formula
The formula itself is straightforward:
Training ROI (%) = [(Value of Outcomes - Cost of Training) / Cost of Training] x 100
A worked example with simple numbers:
- You spend $15,000 on a leadership coaching program for 10 managers
- Over 6 months, those managers’ teams show 4 fewer departures than average
- Each departure costs roughly $45,000 to replace (recruiting, onboarding, lost productivity)
- Value of outcomes: 4 x $45,000 = $180,000
- Training ROI: ($180,000 minus $15,000) / $15,000 x 100 = 1,100%
The formula is the easy part. The hard part is agreeing on the dollar value of outcomes. Turnover cost is the most commonly used because it’s relatively easy to quantify. Other outcomes (engagement, time-to-productivity, customer scores) are real but harder to put a dollar figure on.
When you can’t assign a dollar value, use Tier 2. Behavior change data is credible on its own. “Managers improved 26% on feedback skills” doesn’t require a financial translation. For a side-by-side look at what different training solutions cost, see our pricing comparison guide.
What Risely Measures (And Why It Matters for Your ROI Case)
Most training platforms track completion. Risely tracks three things that map directly to training ROI measurement.
Skill-level improvement across 83 skills. Every user gets baseline assessments when they start. As they work with Merlin (Risely’s AI coach), scores update continuously. You get a clear before-and-after picture at the individual and team level. Across 4,000+ users in 40+ organizations, the average improvement is 26% in 12 weeks.
Behavioral consistency. A single good conversation after training doesn’t count. We track whether managers apply skills repeatedly over time, not just once. This is the difference between “they learned it” and “they actually do it.”
Engagement depth. Not just “did they log in” but how they engage. Are they applying coaching suggestions? Coming back for help with specific situations? An 87% week-1 engagement rate tells you people find it useful enough to return, a much stronger signal than completion.
These three data points give you a Tier 2 measurement framework out of the box. No custom analytics project, no waiting for annual 360 reviews.
For teams evaluating leadership development solutions, built-in measurement changes the ROI conversation entirely. You go from “we think it worked” to “skills improved X% for Y people over Z weeks.” Pricing starts at $59 per user per month, or $399 for a 5-user team.
Build Your Training ROI Case Now
The biggest mistake L&D leaders make with training ROI isn’t using the wrong formula. It’s waiting until the budget review to start measuring. By then, you’re reverse-engineering a justification instead of reporting real results.
Start before the program does. Define success. Pick a tier. Measure at 30, 60, and 90 days.
Need a shareable business case for your next budget conversation? Build one with Merlin. It generates a custom ROI projection based on your team size, industry, and skills. Takes five minutes.
Want to see how the coaching and measurement work before making a case internally? Try Merlin free.
Frequently Asked Questions
What is a good training ROI percentage?
Anything above 100% means you got more value than you spent. Well-designed leadership development programs typically show 150% to 700% ROI when measured through retention impact. But the percentage depends entirely on how you calculate outcome value. A conservative 200% ROI with clear methodology is more credible than a 2,000% ROI with shaky assumptions.
Can you measure training ROI without an LMS?
Yes. An LMS tracks completion and course progress (Level 1 and 2 metrics). Training ROI depends on Level 3 (behavior change) and Level 4 (business results), neither of which require an LMS. You can measure behavior change through 360-degree feedback, manager observations, or skill-tracking platforms. Business results come from HRIS data, turnover reports, and engagement surveys. The LMS is useful for administration, not measurement.
How long does it take to see ROI from training?
Leading indicators (engagement, early skill improvement) show up within 2 to 4 weeks. Behavior change becomes measurable at 8 to 12 weeks. Business results (retention, performance ratings) take 3 to 12 months. This is why 30, 60, and 90-day checkpoints matter. You need early signals to report while you wait for longer-term outcomes.
What is the difference between training ROI and training effectiveness?
Training effectiveness asks “did people learn and apply the skills?” (Kirkpatrick Levels 1 through 3). Training ROI asks “did the financial value of outcomes exceed program cost?” and requires converting effectiveness data into dollar amounts. You can have high effectiveness with unclear ROI if you can’t prove it moved a financial metric. For most L&D teams, effectiveness is the more honest and actionable measure.
How do AI coaching platforms improve ROI measurement?
Traditional training creates a measurement gap: learning happens in a classroom, application happens at the desk, and nothing tracks the transfer. AI coaching platforms like Risely close that gap because the coaching happens where the work happens. Every interaction generates data on which skills are being practiced, how often, and whether improvement sustains. This replaces annual surveys with continuous measurement that L&D leaders can report quarterly.