Most managers either overtrain (hoping people will magically improve) or never train (assuming people arrive perfect). Both are expensive. Here's the economic framework.
Training has value when the cost of building a skill is lower than the cost of buying it. The problem is most managers calculate neither side. They default to whatever feels right, which usually means wasting money on training that doesn't work or losing people who could have been developed.
Start with training cost. It's not just the course fee. Calculate the full load: the program, the time spent training instead of producing, your coaching time, and ramp time until the skill shows up in work. For most skills, that's 3-6 months of reduced productivity plus direct expense. If someone makes $100K and you're training them on a critical skill, you're realistically investing $20-30K total.
Now calculate probability it works. Most training fails. The person attends, returns excited, then reverts to old habits within two weeks because the system hasn't changed. Industry data shows roughly 70% of training produces no measurable behavior change. Your expected value isn't the full benefit. It's 30% of the benefit, weighted by probability it sticks.
Add retention probability. If you invest $25K training someone who leaves within 18 months, you've lost everything. Check how long people stay after significant training. If it's under 24 months, your ROI window is too narrow.
Now calculate the hiring alternative. Recruiting someone who already has the skill costs 20-25% of salary plus 3-6 months ramp time. But you get capability now, not in six months after training maybe works.
Example: You have a manager weak at difficult conversations. The training option looks like $5K coaching plus 15 hours of your time plus 3 months reduced effectiveness. Total cost: $15K. Probability it works: 40%. Expected value: $6K.
Hiring alternative: Replace with someone already good at it. Cost: $30K in recruiting and ramp, but immediate capability. If their role impacts a $2M budget or 10-person team, the productivity difference between weak and strong management is worth $100K+ annually.
The break-even: train when (training cost × success probability × retention probability) is less than (hiring cost + opportunity cost of waiting). In practice, training makes sense for high-retention people in teachable skills you can reinforce. Everything else, hire better.
The principle: train people you're confident will stay, in skills that can be taught, where you can reinforce behavior. Otherwise, upgrade the seat.
