Operations Management (OM) designs and runs the systems that turn inputs into goods and services. Every organisation has operations: a factory, a hospital, a bank branch, a delivery network or a cloud platform. This first part covers how operations are designed; Part II moves to planning, quality and supply chains.
1. The operations function
Operations is a transformation process: inputs (materials, labour, capital, information) become outputs (products and services), with feedback to control it. Together with marketing and finance it is one of the three core business functions.
| Concept | Meaning |
|---|---|
| Goods vs services | Services are intangible, produced and consumed together, perishable, and vary with the provider |
| Productivity | Output ÷ input; single-factor (units per labour hour) or multi-factor |
| Competitive priorities | Cost, quality, speed (delivery), flexibility, innovation |
| Order qualifiers and winners | Qualifiers get you considered; winners get the order |
| Trade-offs | Excelling at all priorities at once is rare |
In the real world: Zara competes on speed and flexibility: it designs and delivers small batches in weeks, accepting higher unit cost in return for fewer unsold items. A discount airline competes on cost and sacrifices flexibility. The operations strategy must match the business strategy.
2. Product and process design
New product development moves through idea generation, screening, concept testing, design, prototyping, testing and launch. Good designs consider standardisation, modularity, simplification and design for manufacture. Value analysis and engineering ask whether each feature earns its cost.
Types of production systems
| System | Volume and variety | Layout | Example |
|---|---|---|---|
| Job (project) | One-off, high variety | Fixed position or process | Shipbuilding, custom software |
| Batch | Medium volume and variety | Process | Bakery, pharma batches |
| Mass (assembly) | High volume, low variety | Product (line) | Cars, smartphones |
| Continuous (process) | Very high volume, no variety | Product | Oil refining, cement, steel |
The product–process matrix links the two: low volume with high variety suits job shops; high volume with standard products suits continuous flow. Choosing a mismatched process is costly.
Service design adds the service blueprint (customer actions, front-stage and back-stage activities), the moment of truth (each customer contact), and the choice of customer involvement. Make-to-stock, assemble-to-order and make-to-order are strategies for when the customer order triggers production. The decoupling point is where stock buffers the customer from production.
In the real world: Dell’s direct model built computers to order, while a car maker builds to forecast; Starbucks lets the customer customise at the end of a standard process (postponement). Ford’s moving assembly line is the textbook mass-production case.
3. Capacity planning
Capacity is the maximum output over a period. Design capacity is the theoretical maximum; effective capacity allows for maintenance and breaks; actual output is what is produced. Two ratios:
- Utilisation = Actual output ÷ Design capacity
- Efficiency = Actual output ÷ Effective capacity
Strategy choices: lead (build ahead of demand), lag (wait for demand) or match. Capacity can be added in lumps, giving economies of scale but risking idle capacity. A bottleneck limits the whole system (theory of constraints, Goldratt): improve the bottleneck, not other steps. Decision trees and break-even analysis help compare capacity options.
In the real world: a restaurant’s bottleneck may be the kitchen, not the seats; a data centre’s may be network bandwidth. In services, capacity cannot be stored, so an empty seat or idle hour is lost forever, which is why airlines and hotels manage capacity through pricing.
4. Facility location and layout
Location
Factors: market and customers, raw materials, labour, transport, power, taxes and incentives, infrastructure and risk. Methods: factor rating, break-even analysis of locations, and the centre of gravity (weighted by volume to find a distribution point). Services are placed near customers; heavy industry near raw materials; and modern firms weigh supply chain resilience (not depending on one site).
Layout
| Layout | Arrangement | Best for | Trade-off |
|---|---|---|---|
| Process (functional) | Similar machines together | Low volume, high variety | Flexible but long flows |
| Product (line) | Machines in sequence | High volume | Efficient but rigid; line balancing equalises work at stations |
| Fixed position | Product stays, resources come | Ships, buildings | Movement of resources |
| Cellular (group technology) | Cells for part families | Medium volume | Mix of both |
| Retail and office layouts | Flow of customers, adjacency | Stores, offices | Sales and interaction |
Line balancing example: if cycle time is set by demand at 2 minutes per unit and total work content is 8 minutes, the minimum number of stations is 8 ÷ 2 = 4.
In the real world: Amazon’s fulfilment centres use chaotic storage and robots to bring items to pickers, while supermarkets place staples at the back so shoppers walk past other goods. Toyota’s cell layouts support its lean system.
5. Forecasting
Operations plans from forecasts. Methods:
| Method | Idea | Use |
|---|---|---|
| Qualitative | Expert opinion, Delphi, market surveys | New products |
| Naive | Next period = last period | Quick baseline |
| Moving average | Average of the last n periods | Stable demand |
| Weighted moving average | Recent periods get more weight | Slightly changing demand |
| Exponential smoothing | F(t+1) = α·A(t) + (1 − α)·F(t) | Recent data matters more; α between 0 and 1 |
| Trend and seasonality | Regression, seasonal indices | Growing or seasonal items |
| Causal | Regression on drivers | Price, promotion effects |
Worked example. Sales for three months were 100, 110 and 120. The 3-month moving average forecast for next month is (100 + 110 + 120) ÷ 3 = 110. With exponential smoothing at α = 0.3, last forecast 105 and actual 120, the next forecast is 0.3 × 120 + 0.7 × 105 = 109.5.
Measure accuracy with MAD (mean absolute deviation), MSE and MAPE. Forecasts are always wrong to some degree, and aggregate forecasts are more accurate than item-level ones, so plan buffers.
In the real world: retailers use sales history, weather and festival calendars to forecast; a wrong forecast means stock-outs (lost sales) or overstock (markdowns). Cloud teams run the same logic for capacity planning and autoscaling.
6. Work study and inventory basics
Work study has two parts: method study (find the best way, using process charts and flow diagrams, and questioning each step: what, why, where, when, who, how) and work measurement (set the standard time by stopwatch time study, work sampling or predetermined times, adding allowances). Ergonomics fits the job to the person.
Inventory buffers uncertainty but costs money (holding, ordering, shortage). Key tools:
| Tool | Idea |
|---|---|
| EOQ | Order quantity √(2DS/H) that minimises ordering plus holding cost |
| ABC analysis | A items (few, high value) get tight control, C items simple control |
| Reorder point | Demand during lead time + safety stock |
| Safety stock | Buffer against variability in demand or lead time |
| JIT and kanban | Pull materials only as needed, cutting stock and waste |
Worked example: D = 2,400 units a year, S = ₹200 per order, H = ₹12 per unit per year gives EOQ = √(2 × 2400 × 200 ÷ 12) = √80,000 ≈ 283 units.
In the real world: Toyota’s just-in-time system relies on small lots and kanban cards, while Indian dairy cooperatives such as Amul manage perishable inventory by daily collection and fast distribution. Lean companies treat excess stock as hiding problems.
One-minute recap
| Topic | Key idea | Where you meet it |
|---|---|---|
| Operations function | Transformation process, productivity, priorities | Strategy, KPIs |
| Product and process design | Job, batch, mass, continuous; blueprint | Factories, service journeys |
| Capacity | Utilisation, efficiency, bottleneck | Plant sizing, autoscaling |
| Location and layout | Factor rating, process vs product layout | Warehouses, stores |
| Forecasting | Moving average, exponential smoothing, MAD | Demand and stock planning |
| Work study | Method study, time study | Productivity, standards |
| Inventory | EOQ, ABC, JIT | Procurement, retail |
Next in MBA Revisited: Marketing Management – I.
Written by Claude, Anthropic’s AI assistant, as part of MBA Revisited, a quick refresher of the PGDM subjects one at a time.