Training return on investment is easier to understand by doing one calculation slowly than by reading definitions. So here is one programme followed from start to finish. A company runs an escalation-handling programme for 60 customer-support team leads. The aim is simple: fewer cases should need to be escalated to senior staff. We will work out what it cost, what it returned and what the answer does and does not tell us.
The framework behind the sums
Two names come up in almost every discussion of training evaluation. Donald Kirkpatrick proposed four levels: reaction (did participants like it), learning (did they pick up the knowledge or skill), behaviour (are they doing things differently at work) and results (did the business metric move). Jack Phillips later added a fifth level, ROI, which converts the results into money and compares them with cost. The formula is simple: ROI percent equals net programme benefits divided by programme costs, multiplied by 100, where net benefits are benefits minus costs.
Most companies stop at level 1 because feedback forms are easy. The interesting work begins when you ask what the training was supposed to change before it started.
Step 1: Add up the full cost
The cost most often forgotten is the participants’ time. In our example:
- Trainer and programme fees: ₹5,00,000
- Participants’ time: 60 leads for 2 days each, at a loaded cost of ₹3,000 a day, which is ₹3,60,000
- Venue, materials and refreshments: ₹40,000
The total is ₹9,00,000. If you had counted only the fee, you would have understated the cost by nearly half.
Step 2: Find the business measure that should move
Before the programme, the company checked its data: about 120 cases a month were escalated beyond the team lead. Each escalation takes up senior staff time, and finance estimates the handling cost at ₹4,000 per case. After the programme, three months later, escalations settled at about 102 a month, a fall of 18.
The annual value of this change is 18 cases times ₹4,000 times 12 months, which comes to ₹8,64,000.
Step 3: Isolate the effect of the training
Here is the step that separates a serious calculation from a flattering one. Escalations did not fall only because of training. In the same quarter, the company improved its knowledge base and the product had fewer bugs. Phillips’s method asks you to estimate what share of the improvement is truly due to the training, using a control group where possible, trend-line analysis, or estimates from participants and their managers, adjusted for their confidence in the estimate.
Suppose the best evidence, from a comparison with a group of leads who had not yet attended, points to about 60% of the fall being due to training. The attributable benefit is 60% of ₹8,64,000, which is ₹5,18,400 a year.
Step 4: Calculate the ROI
Net benefit is ₹5,18,400 minus ₹9,00,000, which is negative ₹3,81,600. Divided by the cost of ₹9,00,000, this gives an ROI of about negative 42 percent in the first year.
Is the programme a failure? Not necessarily, and this is where judgement comes in. The skill does not vanish after a year. If the benefit continues at the same level for a second year with no new cost, the two-year benefit is ₹10,36,800 against a cost of ₹9,00,000, which gives a positive ROI of about 15 percent. Whether to count the second year is a choice, and an honest report states it. Many practitioners prefer to count only the first year or a stated period, to stay conservative. There is also a benefit that we did not count: senior staff time freed for other work, and perhaps better customer satisfaction.
What the number cannot tell you
- The 60% attribution is an estimate. Change it to 40% or 80% and the ROI moves a lot. Show the sensitivity instead of hiding it.
- Not every programme has a hard metric. Leadership or communication training rarely maps cleanly to a rupee figure, and forcing one invites invention. For such programmes, evidence at levels 2 and 3, such as observed behaviour and feedback from the team, is more honest than a made-up ROI.
- A good ROI says nothing about whether this was the best use of ₹9 lakh. Fixing the knowledge base may have returned more for less.
- Short programmes with an easily measured aim, such as a product training for sales, are the best candidates. Apply the full method to a few expensive, strategic programmes, and use lighter checks for the rest.
The useful habit is not to chase a precise percentage. It is to decide, before the programme starts, which behaviour and which business measure should change, to record the baseline, and to look again three months later. Even a rough answer from that discipline is better than a thick folder of happy feedback forms.
Figures and names in the examples are illustrative.
Leave a Reply