A back office processed 1,20,000 transactions with more hours and a higher hourly rate than standard. Split the labour overrun into rate and efficiency variances, and say which matters more.

The overrun is Rs 40.2 lakh: Rs 13.2 lakh adverse on rate and Rs 27.0 lakh adverse on efficiency. Standard hours for 1,20,000 transactions are 60,000, so staff used 6,000 hours, 10%, more than they should. Paying Rs 20 more on 66,000 hours adds the rest. Efficiency matters more: it is twice the size, and how long work takes is more within the manager's control than the market wage.

The standard hours for the output actually produced. Variances compare actual cost with what the actual volume should have cost, not with the original budget: 1,20,000 transactions at 0.5 hours is 60,000 standard hours, Rs 270 lakh at Rs 450. Actual cost was 66,000 hours at Rs 470, Rs 310.2 lakh. The gap of Rs 40.2 lakh is what you now split, like a grocery bill that is higher because you bought more and because prices rose.

Change one thing at a time. The rate varianceThe part of a cost overrun caused by paying a different price per unit of input than standard, measured on the inputs actually used. prices the actual hours at the difference in rate; the efficiency variance prices the extra hours at the standard rate. (470 less 450) times 66,000 is Rs 13.2 lakh. (66,000 less 60,000) times 450 is Rs 27.0 lakh. The two add back to Rs 40.2 lakh exactly, which is the check that you used the right hours in each formula.

Efficiency, for two reasons. It is the larger number, and it is the more controllable one: hours per transaction depend on process, training, system downtime and rework, all things an operations head can act on, while the hourly rate is largely set by the labour market and the annual increment. A 10% efficiency slip on a process that should be stable is a signal; a 4.4% rate rise may simply mean the standard is out of date.

Then check whether the two are linked. If the extra hours were worked as overtime at a premium, part of the rate variance is a consequence of the efficiency problem, not a separate one. Note too that the rate variance on actual hours contains a small joint piece: Rs 20 on the 6,000 extra hours, Rs 1.2 lakh, which some teams report on its own. Say which convention you used. The judgement to close on: fix the hours first, and ask whether the Rs 450 standard should be reset before calling the rate a failure.