Labour Productivity Analysis for SMEs

Labour cost is a major share of operating expense in most manufacturing operations, and yet labour productivity is one of the least-measured dimensions of performance. Most managers know labour cost in rupees per month — far fewer know labour productivity in output per hour or output per rupee. Here is the working method to track it without overcomplication.

The basic metrics

Three foundational labour productivity metrics:

• Output per labour hour = Units produced ÷ Direct labour hours worked. The cleanest physical productivity measure.
• Labour cost per unit = Total direct labour cost ÷ Units produced. The financial productivity measure.
• Labour efficiency = Standard hours for actual output ÷ Actual hours worked. The variance-driven measure.

Each tells you something different and they complement each other.

Output per hour — the operational metric

This is the easiest to start with. For each production process:

• Standard ideal output per hour (from time studies, BOM standards, or historical best)
• Actual output per hour by shift, by operator, by week
• Trend over time

If you make 50 cabinets per day with 4 operators working 8 hours each = 32 hours, output per hour = 1.56 cabinets/hour. If the standard is 2 cabinets/hour, you're at 78% of standard productivity.

Track this weekly. A drop of 5% sustained for 2 weeks is a flag worth investigating.

Labour cost per unit — the financial metric

Labour cost per unit captures both productivity and labour rate effects:

• Labour cost = labour hours × hourly cost
• Labour cost per unit = (Labour hours × Hourly cost) ÷ Units produced

A rising labour cost per unit could mean:

• Lower productivity (more hours per unit) — operational issue
• Higher labour rates (overtime, mix shift to senior workers) — HR / mix issue
• Lower output (volume down, fixed labour spread over fewer units) — sales / planning issue

Just like cost per unit overall (see our cost per unit trends post), the decomposition matters more than the headline change.

Labour efficiency — the variance lens

Labour efficiency variance = (Standard hours for actual output − Actual hours worked) × Standard rate

If you made 50 cabinets that should take 3 hours each (standard hours = 150) but the team spent 175 hours, efficiency variance = (150 − 175) × ₹250 = ₹6,250 unfavourable.

That's ₹6,250 of labour cost spent in excess of standard for the same output. Aggregated over the month, it's the size of the labour efficiency gap.

See our variance analysis post for the full mechanics.

Where productivity loss hides

Labour hours that don't produce output:

• Wait time — waiting for material, machine, instructions
• Setup / changeover time — not always tracked separately from production time
• Rework — fixing units that should have been right first time
• Meetings and admin — necessary but should be bounded
• Material handling — moving things rather than producing
• Quality checks — necessary but often longer than ideal

A simple time study — once or twice a year, for one or two shifts — that breaks an 8-hour shift into producing time, setup time, wait time, rework time, and other shows where the leverage is.

A typical SME shop floor with no measurement starts at 50-60% "value-added time" (actual production); world-class is 80%+. The gap is the prize.

Productivity by what dimension?

Aggregate productivity is fine for trend; per-dimension productivity is where action happens:

• Per shift — first shift vs second; different supervision, different team
• Per operator — individual differences; training gaps; outstanding performers to learn from
• Per product — some products are intrinsically slower; per-product standards
• Per workstation / line — where the bottleneck is
• Per day of week — Monday is often the worst day; understand why

Variance across dimensions is the diagnostic data.

What good labour productivity looks like

Difficult to benchmark across industries — too many variables. The right benchmarks:

• Your own trend — improving or declining over months
• Your best operators / shifts — if your best team operates at 85% of standard, your aggregate target should be 80%, not 60%
• Industry benchmarks — published or trade-association data, where available
• Pre-and-post a specific change — bringing in a new machine, restructuring a line, training programme

The relative comparison is more reliable than the absolute one.

Driving labour productivity — what actually works

• Time standards and visible tracking — operators who can see their output vs target perform better than those who can't
• Reduce setup / changeover time (the SMED approach) — often the biggest single labour productivity lever
• Training in the actual work, not in classrooms — on-the-job coaching from the best operators
• Eliminate wait time — make sure material arrives, machines are ready, instructions are clear before the shift starts
• Multi-skill operators — flexibility reduces wait time and balances load
• Pay incentives tied to productivity — but designed carefully to not encourage quality compromise

Each of these is a project in itself. The measurement is what tells you which to prioritise.

The pay vs productivity question

A frequent question: should we pay more to attract better workers, or push productivity with the workers we have?

The answer is almost always both, in sequence:

• First, measure so you know what productivity actually is
• Second, structure work to make current workers more productive (training, methods, tools, wait-time elimination)
• Third, pay competitively for the productive workers you've now identified
• Fourth, selectively upgrade with new hires when natural turnover happens

Going straight to "pay more" without changing the work usually produces marginal improvement at higher cost. Measurement and structure first; pay adjustment after.

The cultural dimension

Labour productivity measurement is sensitive. Workers can react badly if it feels like surveillance ("are you watching me work?") rather than support ("we're trying to make your shift productive").

Communication matters: the data is for finding obstacles to performance, not for tracking individuals. The first action on a productivity drop should be a question — "what's blocking us?" — not an accusation. Teams that experience measurement this way embrace it.

How Booksmor helps

Booksmor captures direct labour against production runs and computes output per hour, labour cost per unit, and labour efficiency variance — per shift, per operator, per product, per work centre. Trend dashboards and exception flags surface productivity issues weekly rather than at year-end. Start a 30-day free trial and put labour on the dashboard.