Training is one of the few spend categories with a measurable, compounding return, and skipping it doesn't save money. it just moves the cost somewhere less visible.
Untrained staff don't cost nothing. They cost you in:
Error rates - mistakes that need rework, refunds, or client apologies
Ramp time - new hires taking months longer to reach full productivity
Turnover - people leave roles they feel unequipped to do well, and replacing them costs 6–9 months' salary on average
Tribal knowledge risk - one person holds the process in their head; they leave, and it leaves with them
Where Training Actually Fails
Most training doesn't fail because the content was bad. It fails because of what happens (or doesn't) around it:
No defined owner - someone runs the session, then nobody checks if it changed behaviour
Generic content, specific gap - an off-the-shelf course that doesn't map to the actual process breaking down
No follow-through - a one-off event instead of reinforcement, so the lift decays within weeks
Measured by attendance, not outcome - "12 people completed the module" tells you nothing about whether errors dropped
If you can't name what changed after the training, you didn't train ,you ran an event.
Training as a Systems Lever
Treat your business as a system, not a set of silos. A gap in one function say, document control shows up as delay in another (project delivery) and cost in a third (rework, client trust). Training the person closest to the gap is often the cheapest fix available, cheaper than new software, cheaper than hiring, cheaper than absorbing the error rate indefinitely.
The question isn't "should we train?" It's "which gap, closed now, saves us the most downstream?"
