Agribusiness Management · Agri-sector specialisation · also ABM 508, Logistics & SCM (MANAGE), Food Supply Chain Management (IIMA elective), Sustainable Supply Chain Analytics (WUR)
Moving agricultural and food products efficiently from source to buyer: planning, procurement, inventory, transport, warehousing, cold chain and outsourced logistics.
In the exams: agribusiness management PhD entrance and fellowship examinations, logistics and supply chain management (heavy)
Agri supply chain and logistics management is the study of how agricultural inputs and food products move from where they are made to where they are used, and of how the firms along the way plan, buy, store, move and account for them. A supply chain is the set of organisations and activities, from input suppliers through farmers, processors and retailers, that turn raw materials into a product in a buyer's hands. Logistics is the part of supply-chain management that plans and controls the forward and reverse flow and storage of goods, services and related information between the point of origin and the point of consumption, so as to meet customer requirements. Supply-chain management is wider: it also covers sourcing decisions, relationships between firms, demand planning and the coordination of all of these.
Agricultural chains are harder than manufacturing chains. Supply is seasonal and weather-dependent while demand is year-round; products are perishable, so time and temperature destroy value; raw produce is bulky and cheap per tonne, so transport and storage are a large share of price; quality varies by lot; many chains start with millions of smallholders; food rules require tracing to source; and post-harvest loss and consumer waste are both economic and food-security problems.
The subject is examined at doctoral entrance in agribusiness (in India it is a heavily weighted part of the operations unit of the ICAR PhD entrance for Agri-Business Management), in food and agribusiness master's courses worldwide, and in operations-research papers, where inventory models are set as numericals. This manual applies the standard supply-chain and logistics textbooks to grain, fresh produce, dairy, meat and inputs. Every numerical example was computed by script; figures are illustrative, and prices, tariffs and rules vary by country and year.
An agri-food supply chain links input supply (seed, fertiliser, crop protection, feed, veterinary products, machinery), production on farms, aggregation, primary processing (cleaning, grading, drying, milling, chilling, slaughter), secondary processing (manufactured foods), distribution (wholesale, retail, food service, export) and consumption. Three flows run through it. The product flow moves downstream from farm to consumer, with a smaller reverse flow of returns, packaging and waste. Money flows upstream. Information flows both ways: orders and demand signals go up, and specifications, prices and shipment notices go down.
A supply chain view analyses the physical and transactional links for cost and service; a value chain view, from Porter's work and later development economics, asks where value is added, who captures it and how the chain is governed.
The features of agricultural chains listed in the introduction each have a logistics consequence: seasonality means someone must hold stock; uncertain yields mean supply must be forecast as well as demand; perishability and variable quality mean grading, cooling and speed matter; low value per tonne makes transport costly relative to value; dispersed small farms make aggregation the costly first step; and food regulation makes records and traceability part of the chain.
A traditional agricultural marketing chain, common in South Asia, sub-Saharan Africa and parts of Latin America and Southeast Asia, passes produce from the farmer through several intermediaries: a village trader or assembler, a commission agent in a primary or regulated market, a wholesaler in a terminal market, and a retailer, often a small shop, street vendor or open market stall. In India, regulated markets under state APMC laws are the central node.
Traditional chains are flexible, buy any quantity and quality, often lend to farmers and need little capital. Their weaknesses are many handling stages, each adding cost and loss; little grading, so quality is not rewarded; poor storage and non-refrigerated transport; weak price information for farmers; a low farmer's share of the consumer price for many perishables; and, because transactions are spot and anonymous, little feedback about consumer wants and little traceability.
Modern agri-food chains are shorter and more coordinated: direct procurement through buyers' collection centres, contract farming, aggregation by farmer producer organisations (FPOs) and cooperatives, private quality and safety standards, pack-houses that grade and pre-cool near the farm, distribution centres, refrigerated transport and sales data shared with suppliers. Reardon and colleagues described the rapid spread of supermarkets in developing countries from the 1990s as a "supermarket revolution", first in processed foods and later in fresh produce, and showed that it pushed procurement towards specialised wholesalers, preferred suppliers and private standards.
The drivers are rising urban incomes, demand for convenience, safety and year-round supply, investment in retail and processing, cheaper information technology, and reforms allowing private markets and direct purchase. Electronic platforms, such as India's e-NAM linking regulated markets, and online grocery add channels.
Buyers in modern chains want volume, uniform quality and certification, which favours larger or organised farmers, and buyer power can shift risk onto farmers; policy responses include support for cooperatives and FPOs, public pack-houses and cold stores, and contract regulation. Most countries run both kinds of chain side by side, with modern chains growing fastest in high-value produce, dairy, poultry and processed foods.
Chopra and Meindl frame supply-chain strategy around strategic fit: a firm's supply chain must match the uncertainty its customers and suppliers impose. Fisher's classic distinction, published in 1997, divides products into functional products, with stable and predictable demand, long life cycles and low margins (staple grains, flour, milk, sugar), and innovative products, with unpredictable demand, short life cycles and high margins (new branded snacks, fashion-driven foods). Functional products need an efficient supply chain that minimises cost through high utilisation, low inventory and cheap transport. Innovative products need a responsive supply chain that holds buffer capacity and stock and uses faster transport, because the cost of a lost sale or of unsold obsolete stock is high. Many agri chains are efficient-type for bulk commodities and responsive-type for fresh and branded products, and some firms run both.
Chopra and Meindl name six drivers that set the balance between efficiency and responsiveness. The logistical drivers are facilities (where plants, warehouses and collection centres are, how big and how flexible), inventory (how much raw material, work-in-process and finished stock is held and where) and transportation (which mode, route and frequency). The cross-functional drivers are information, sourcing (in-house or outsourced) and pricing (used to shape demand). Fewer, larger warehouses are the efficient setting; many small ones near customers are the responsive setting.
Service metrics: order fill rate, unit fill rate (share of units supplied from stock), cycle service level (probability of no stockout in a cycle), on-time-in-full (OTIF) and the perfect order rate. A perfect order is delivered complete, on time, undamaged and with correct documents; if these are independent, the perfect order rate is the product of the component rates, so four components at 95, 98, 97 and 99 percent give only about 89 percent perfect orders.
Inventory metrics: average inventory, inventory turnover (cost of goods sold divided by average inventory at cost) and days of inventory (365 divided by turns, or average inventory divided by daily cost of goods sold). Little's law links them: average inventory equals throughput rate multiplied by average flow time, so a plant that processes 400 tonnes a day and holds 6,000 tonnes in the system has an average flow time of 15 days.
Cost and financial metrics: total logistics cost as a share of sales, cost per tonne-kilometre, and the cash-to-cash cycle, which equals days of inventory plus days of receivables minus days of payables (45 + 30 - 20 = 55 days, for example). Agri-specific metrics include post-harvest loss, cold-chain temperature compliance and utilisation of seasonal assets such as silos.
The SCOR (Supply Chain Operations Reference) model organises processes into plan, source, make, deliver and return, with enable processes, and ties metrics to reliability, responsiveness, agility, cost and assets; later versions renamed the process groups.
The main obstacles to coordination are incentive conflicts (each stage optimises its own profit), information distortion, operational practices such as large batches and long lead times, pricing practices such as trade promotions, and behavioural factors such as poor trust. The best-known symptom is the bullwhip effect: the variability of orders increases as one moves upstream from retailer to wholesaler to processor to farmer, even when consumer demand is fairly stable. It was named and analysed by Lee, Padmanabhan and Whang in 1997, who gave four causes: demand signal processing (each stage updates its forecast from the orders it receives rather than from end demand, and adds safety stock, amplifying swings); order batching (ordering in large lots weekly or monthly for transport or ordering economies); price fluctuation (promotions cause forward buying followed by no orders); and rationing and shortage gaming (when supply is short and allocated in proportion to orders, buyers inflate orders and cancel later).
The bullwhip effect is often measured as the ratio of the variance of orders placed by a stage to the variance of the demand it receives; a ratio above one indicates amplification. Order variance of 900 against demand variance of 100 gives a ratio of 9. The remedies follow from the causes: share point-of-sale data and forecasts upstream, reduce lead times, order in smaller and more frequent lots (made cheaper by mixed loads and third-party carriers), stabilise prices (everyday low pricing instead of deep promotions), allocate scarce supply on past sales rather than current orders, and use vendor-managed inventory or collaborative planning.
Every supply-chain decision rests on a forecast, from a dealer's seed stock before sowing to a dairy's flush-season intake. Chopra and Meindl list four characteristics: forecasts are always inaccurate, so they should carry an expected value and a measure of error; long-term forecasts are less accurate than short-term ones; aggregate forecasts (for a product family or a region) are more accurate than disaggregate ones (for one pack size in one store); and the further up the chain a firm is from the consumer, the more distorted the information it receives, which is the bullwhip effect again.
Agricultural forecasting has two sides. Demand forecasting estimates what buyers will take. Supply forecasting estimates what farms will deliver: crop area, yield and harvest timing, or milk and egg output. For a processor, supply is often the harder forecast, built from sown-area surveys, crop reports, weather, remote sensing and field staff. Demand planning turns forecasts into an agreed plan, often through a monthly sales and operations planning (S&OP) meeting that reconciles sales with procurement, production and logistics capacity.
Qualitative methods, used when data are scarce, include the Delphi method (anonymous rounds of expert opinion converging to consensus), jury of executive opinion, sales-force composite and market surveys. Time-series methods use the history of demand alone and assume the pattern will continue. Causal methods relate demand to other variables, such as price, income, promotions, rainfall or festivals, usually by regression.
A time series is decomposed into level, trend, seasonality and a random component; festival and summer peaks give agri demand strong seasonality. The naive forecast uses last period's actual value. A simple moving average forecasts the next period as the mean of the last n periods; with sales of 420, 460 and 500 tonnes in the last three months, a three-month moving average forecasts 460 tonnes. A weighted moving average gives more weight to recent periods. Simple exponential smoothing updates the forecast by a fraction alpha of the latest error: new forecast = old forecast + alpha x (actual - old forecast). If last month's forecast was 1,000 tonnes, actual sales were 1,200 and alpha is 0.3, the new forecast is 1,000 + 0.3 x 200 = 1,060 tonnes. A higher alpha responds faster but also chases noise. Holt's method adds a smoothed trend, and the Holt-Winters method adds seasonal factors. A seasonal index, the ratio of a season's average to the overall average, deseasonalises data before a trend is fitted and reseasonalises the forecast afterwards.
Forecast error for a period is actual minus forecast. The main measures are the mean absolute deviation (MAD), the average of the absolute errors; the mean squared error (MSE), which penalises large errors more; the mean absolute percentage error (MAPE), the average of absolute errors divided by actual demand, which allows comparison across products of different sizes; and bias, the sum or mean of the signed errors, which shows whether a method consistently over- or under-forecasts. For actual demand of 100, 120, 90 and 110 against forecasts of 110, 110, 100 and 100, the errors are -10, +10, -10 and +10, so the MAD is 10, the bias is zero and the MAPE is about 9.6 percent.
The tracking signal is the cumulative forecast error divided by the MAD. When it drifts outside chosen limits, commonly plus or minus 4 in textbook examples (firms set their own limits), the forecast is judged biased and the method is reviewed. The standard deviation of forecast error, estimated as roughly 1.25 times the MAD when errors are normally distributed, is what feeds safety-stock calculations, which is why forecasting and inventory management are taught together.
Procurement is obtaining goods and services, from identifying a need to paying the supplier; purchasing is its transactional core. Agribusiness buys in two different markets, industrial inputs and packaging, and farm produce from many farmers, traders or cooperatives; raw material is usually a processor's largest cost.
The purchasing cycle runs: need recognition and specification (a purchase requisition); supplier identification; request for quotation; evaluation and selection on price, quality, reliability, capacity and financial soundness; negotiation; purchase order; follow-up and expediting; receipt, weighing and quality inspection; the three-way match of order, goods receipt and invoice before payment; and supplier performance evaluation.
The Kraljic matrix classifies purchases by profit impact and supply risk into four groups: strategic items (high impact, high risk), which need long-term partnerships; leverage items (high impact, low risk), where buying power and competitive bidding pay; bottleneck items (low impact, high risk), where security of supply matters most; and non-critical or routine items (low impact, low risk), where the aim is to simplify and cut transaction cost. For a tomato processor the crop is strategic, cans leverage, a single-source enzyme a bottleneck.
A buyer of farm produce can use the spot market, forward agreements, contract farming, cooperatives and FPOs, or its own farms. Transaction-cost reasoning guides the choice: the more specific the investment, the more quality and timing matter and the less reliable the spot market, the further the buyer moves towards contracts or integration.
Specifications cover variety, grade, moisture, foreign matter, damage, residues and microbiology. Milk is commonly paid on fat and solids-not-fat content; grain is priced against a base grade with discounts for moisture, foreign matter and broken or damaged kernels. Weight adjustment for moisture is a common numerical: if 100 tonnes of grain at 18 percent moisture is dried to 13 percent, the dry matter (82 tonnes) is unchanged, so the new weight is 82 / 0.87, about 94.25 tonnes, a shrink of about 5.75 tonnes rather than the 5 tonnes a naive subtraction of percentage points suggests.
Contract farming is agricultural production carried out under an agreement, made before production, between a buyer and farmers that sets conditions for the production and marketing of the product. Mighell and Jones distinguished three types by what the contract covers: market-specification contracts (the buyer promises a market, often with a price formula, for a product meeting a specification), resource-providing contracts (the buyer also supplies inputs such as seed, chicks, feed or credit, recovered from the payment) and production-management contracts (the buyer also specifies and supervises production practices). Eaton and Shepherd's FAO guide describes five organisational models: the centralised model (a processor contracts many farmers, with tight quality control, as in sugar cane, tobacco, poultry and processing vegetables); the nucleus estate model (the company runs its own estate or plantation and contracts surrounding outgrowers); the multipartite model (several parties, such as a company, a government agency and a farmer organisation, share roles); the informal model (small firms or traders with simple, often seasonal, verbal agreements); and the intermediary model (the company contracts through a middleman or collector who contracts farmers).
The buyer gains volume, quality, timing and traceability without owning land; farmers can gain a market, inputs, credit, technology and less price risk. The known problems are side-selling by farmers when the spot price rises above the contract price, rejection of produce by the buyer on quality grounds when the market price falls, delayed payment, imbalance of bargaining power, and disputes over grading. Good contracts specify quality testing, pricing (fixed, market-linked formula, or a floor with shared upside), input recovery, delivery schedules and dispute resolution. Countries regulate contract farming in different ways; in India a model state law was circulated in 2018, and a central act of 2020 on farming agreements was repealed in 2021, so the legal position depends on the state. The UNIDROIT/FAO/IFAD Legal Guide on Contract Farming (2015) is the international reference.
Corporate farming means a company producing crops or livestock directly on land it owns or leases, using hired labour and management. It gives the most control but ties capital to land and carries production risk; land ceiling, tenancy and ownership laws restrict it in some countries, while plantations and large grain and livestock operations are common in others. Firms often combine own, contracted and spot supply.
Agribusinesses hold cycle stock, because ordering or transporting in lots is cheaper than continuous small deliveries; safety (buffer) stock, against uncertainty in demand and lead time; seasonal or anticipation stock, because harvest comes once or twice a year while use is continuous; pipeline or in-transit stock, which is goods moving between stages; and speculative stock, held because a price rise is expected. Seasonal stock dominates in grain, much of it held by governments for food security. A trader stores from harvest only if the expected price rise over the storage period exceeds the full cost of carrying the stock (interest, storage charges, shrink and quality loss); when it does not, storing is a loss even if prices rise.
Ordering (set-up) cost is incurred per order. Holding cost is incurred per unit per period: capital tied up (usually the largest part), space, insurance, handling, and deterioration and shrinkage, which are large for perishables. Shortage cost is the lost margin, customer or emergency purchase.
The economic order quantity (EOQ, Harris-Wilson lot size) minimises the sum of annual ordering cost and annual holding cost. With annual demand D, cost per order S and holding cost per unit per year H, annual ordering cost is (D / Q) x S and annual holding cost is (Q / 2) x H, because average cycle stock is half the order size. Setting their derivative to zero gives Q* = square root of (2DS / H). At the EOQ, annual ordering cost equals annual holding cost.
Example: a feed mill uses 12,000 tonnes of a premix ingredient a year, each order costs 400 in currency units, and holding costs 6 per tonne per year. Q* = square root of (2 x 12,000 x 400 / 6) = about 1,265 tonnes. The mill places about 9.5 orders a year at a combined ordering and holding cost of about 7,589. A student who drops the factor 2 gets about 894 tonnes, a common slip.
Example with a percentage holding cost: a dealer sells 24,000 bags of fertiliser a year, each order costs 750, a bag costs 1,500 and the annual carrying rate is 20 percent. H = 0.20 x 1,500 = 300 per bag per year, so Q* = square root of (2 x 24,000 x 750 / 300) = about 346 bags. Using 20 instead of 0.20 gives about 35 bags, a factor of ten too small.
Because Q* depends on the square root of demand, four times the demand only doubles the order size. The EOQ assumes constant known demand, a fixed ordering cost, a constant holding cost, instantaneous replenishment, no shortages and no quantity discounts. With quantity discounts, the procedure is to compute the EOQ at each price, raise it to the minimum quantity of the break if it falls below, and compare total annual cost including purchase cost across the candidate quantities. For example, with demand of 10,000 units, ordering cost 500, holding at 20 percent of price, and prices of 50 below 2,000 units and 48 from 2,000 units, the EOQ at 48 is about 1,021 units, which is infeasible for that price, so the candidate is 2,000 units; its total annual cost (about 492,100) is lower than that of the 1,000-unit EOQ at price 50 (510,000), so the firm should order 2,000. The total cost curve is flat near the optimum, so the EOQ is robust to small errors in D, S and H.
In a continuous-review system, the firm orders a fixed quantity (often the EOQ) whenever inventory position falls to the reorder point (ROP). ROP = expected demand during lead time + safety stock. If daily demand is 200 bags, lead time 6 days and safety stock 300 bags, the ROP is 200 x 6 + 300 = 1,500 bags. Inventory position is stock on hand plus on order minus backorders.
When daily demand is normally distributed with standard deviation sd and lead time is L days and fixed, the standard deviation of demand over the lead time is sd x square root of L (variances add across independent days). Safety stock = z x sd x square root of L, where z is the standard normal value for the chosen cycle service level: about 1.28 for 90 percent, 1.645 for 95 percent, 1.96 for 97.5 percent and 2.33 for 99 percent. With daily sd 40 units, lead time 9 days and 95 percent service, safety stock is 1.645 x 40 x 3, about 197 units; multiplying by 9 instead of the square root of 9 gives 592, three times too much.
When lead time also varies, with mean L and standard deviation sL, and daily demand has mean d and standard deviation sd, the standard deviation of lead-time demand is the square root of (L x sd squared + d squared x sL squared). With d = 500, sd = 80, L = 4 days and sL = 1 day, this is about 525 units, and at z = 1.65 the safety stock is about 866 units.
In a periodic-review system, stock is checked every T periods and topped up to an order-up-to level that must cover demand over T + L, so periodic review needs more safety stock than continuous review for the same service.
Service can be defined as cycle service level (probability of no stockout in a cycle) or fill rate (fraction of demand met from stock). A 95 percent cycle service level usually gives a fill rate well above 95 percent, because a stockout affects only part of a cycle's demand.
For a perishable bought once for one selling period, such as festival flowers or a day's bread, the EOQ does not apply. The newsvendor (single-period) model sets the order where the probability of selling all units equals the critical ratio Cu / (Cu + Co), where Cu is the cost of underage (margin lost per unit short, price minus cost) and Co the cost of overage (loss per unit left over, cost minus salvage value). With price 50, cost 20 and salvage 5, Cu = 30 and Co = 15, so the critical ratio is 2/3; for normal demand with mean 1,000 and standard deviation 200, z is about 0.43 and the optimal order is about 1,086. When Cu equals Co, the order equals mean demand. Higher salvage value (a processor who will buy unsold produce) raises the optimal order.
ABC analysis ranks items by annual usage value; the few A items that carry most of the value get tight control. VED (vital, essential, desirable) and FSN (fast, slow, non-moving) are related classifications.
Risk pooling: aggregating demand across locations cuts safety stock because fluctuations partly cancel. When n locations with independent and identical demand are served from one central stock, total safety stock falls by a factor of about the square root of n; this is often called the square-root law of inventory. Centralisation saves inventory but raises outbound transport cost and delivery time.
Independent demand (for end products) is forecast; dependent demand (for materials and components) is calculated from the production plan. Material requirements planning (MRP) handles dependent demand. Its three inputs are the master production schedule (MPS: how many units of each end product are to be finished in each period), the bill of materials (BOM: the list and quantities of materials and components per unit of end product, arranged in levels) and the inventory status file (stock on hand, scheduled receipts and lead times).
MRP explodes the MPS through the BOM to obtain gross requirements for each material in each period, then nets them: net requirement = gross requirement - stock on hand - scheduled receipts + safety stock (where a safety stock is to be maintained). Planned orders are then offset backwards by the lead time. For example, if a juice plant needs 5,000 cartons next week, holds 1,200, has 800 on order due before then and keeps 500 as safety stock, the net requirement is 5,000 - 1,200 - 800 + 500 = 3,500 cartons, to be released one lead time earlier. Lot-sizing rules (lot-for-lot, EOQ, period order quantity) group net requirements into orders.
In food plants MRP covers ingredients, packaging and spares, and the BOM is a recipe or formulation. The main raw material arrives on the farm's timetable, so the plant plans around a supply forecast, and recipes may change with raw-material quality such as milk fat. MRP II added capacity, labour and finance; ERP integrates these with sales, procurement and accounting.
Just-in-time (JIT), from the Toyota Production System, aims to produce and deliver exactly what is needed, when needed, and to remove waste. Its features are small lots through set-up reduction, pull scheduling, kanban signals that authorise production only when downstream stock is used, few suppliers delivering frequently, total quality management and a level schedule. Lean lists seven wastes: overproduction, waiting, transport, over-processing, inventory, motion and defects.
The number of kanban containers is often computed as N = d x L x (1 + alpha) / C, where d is demand rate, L the replenishment lead time, alpha a safety factor and C the container size. With demand of 300 units per day, lead time half a day, a 20 percent safety factor and containers of 40, N = 300 x 0.5 x 1.2 / 40 = 4.5, rounded up to 5 containers.
JIT fits agribusiness unevenly. It works for packaging and ingredients from industrial suppliers, for fresh products delivered daily to retailers (bread, milk, fresh meat), and in integrated poultry, where chick placements are scheduled to match processing slots. It does not fit the main raw material of seasonal crops, which must be harvested within a short window and stored, and it is fragile when transport is unreliable. Low inventory also gives less protection against droughts, disease outbreaks and port closures; after the 2020 pandemic disruptions many firms added buffers, a shift described as from just-in-time to just-in-case.
In vendor-managed inventory (VMI), the supplier monitors the customer's stock, usually through shared electronic data on sales and inventory, and decides when and how much to replenish within agreed minimum and maximum levels. Ownership of the stock may stay with the supplier until it is used or sold (consignment) or pass on delivery; the two are separate choices. VMI reduces the bullwhip effect because the supplier sees real demand rather than batched orders, lets the supplier plan production and transport better, and saves the customer ordering effort. Its risks are the customer's loss of control, overstocking when the supplier does not bear holding cost, and dependence on trust and data.
Collaborative planning, forecasting and replenishment (CPFR) goes further: trading partners share and agree forecasts, promotions and replenishment plans. Efficient consumer response (ECR) in grocery shares the aim.
Logistics is summarised as the seven rights: the right product, quantity, condition, place, time, customer and cost. Its elements are customer service; order processing, which starts all physical movement; inventory management; transportation; warehousing; materials handling; packaging, which protects and unitises product; facility location and network design; return goods handling; and the information system linking them.
The total cost concept holds that these costs must be managed together because they trade off: cheaper, slower transport raises in-transit and safety stock; fewer warehouses cut storage and inventory but raise outbound transport; cheaper packaging raises damage. The aim is the lowest total cost for a given service level, not the lowest cost of each element.
The centre-of-gravity method gives a first location estimate: x is the volume-weighted mean of the points' x coordinates, and likewise y. For three markets at (10, 20) with 300 tonnes, (40, 10) with 100 tonnes and (30, 50) with 200 tonnes, the centre of gravity is at about (21.7, 28.3). It ignores roads and land cost, so factor rating, transport-model linear programmes and network optimisation refine it. For agricultural assets, location also depends on where the crop is grown: weight-losing processes (sugar cane, oil palm fruit, fresh milk) locate near the source, as Weber's location theory predicts.
Chopra and Meindl describe six designs for moving product from manufacturer to customer: manufacturer storage with direct shipping (drop-shipping); manufacturer storage with direct shipping and in-transit merge; distributor storage with carrier delivery; distributor storage with last-mile delivery to the customer's door, as in grocery home delivery; manufacturer or distributor storage with customer pickup at a collection point; and retail storage with customer pickup, the traditional shop. They differ in inventory, transport cost, response time, variety, availability and ease of returns; grocery relies mostly on distributor storage and stores, and online grocery adds costly last-mile delivery.
Between the plant and the store, the main flow options are direct shipment (each supplier ships straight to each store, suited to full truckloads), shipment through a distribution centre where stock is held, cross-docking (inbound goods sorted straight onto outbound vehicles within hours, suiting fast-moving and perishable goods), and milk runs (one vehicle collects from several suppliers or delivers to several customers on one route). Milk collection routes from village centres to a chilling plant are the literal case.
Pool distribution is a way of serving many small customers in a distant region at close to truckload cost. A shipper, or a pool distributor on its behalf, combines many small orders bound for the same region into one full truckload, sends it to a terminal or cross-dock in that region, and there breaks it into local deliveries on smaller vehicles. It replaces many less-than-truckload (LTL) shipments or parcel shipments, each paying high per-unit rates, with one long-haul truckload plus short local runs. The cost is an extra handling stage, and it needs steady volume to the region.
Consolidation is the wider principle. Shipment consolidation can be by time (holding orders for a few days until a full load is reached, which delays service), by place (combining loads for the same region, as in pool distribution), by product (combining several products on one vehicle) or by vendor (a consolidator collecting from several suppliers for one buyer). Less-than-container-load (LCL) sea freight is consolidation sold commercially.
Pallet and crate pools, which rent reusable pallets and crates to growers and packers, are a different use of the word and belong to reverse logistics.
The modes are road, rail, water, pipeline and air, compared on cost, speed, reliability, capability, accessibility and frequency. As a general ranking of cost per tonne-kilometre for bulk freight, water and pipeline are usually cheapest, then rail, then road, with air the most expensive; the ranking for speed is roughly the reverse. Road dominates the first and last mile because it reaches farms directly and is flexible; rail and water win on long-haul bulk; air carries high-value perishables such as cut flowers.
Intermodal transport uses two or more modes with the load in one unit, usually a container, so the goods are not handled when the mode changes. Reefer containers carry most of the world's seaborne perishable trade.
Two principles govern transport cost. Economy of scale: the cost per unit of weight falls as the size of the shipment rises, because fixed costs of loading, paperwork and the vehicle trip are spread over more units, which is why truckload (TL) rates per tonne are lower than less-than-truckload (LTL) rates. Economy of distance: cost per kilometre falls as distance rises, because terminal costs of loading and unloading are spread over more kilometres, which is why freight rates taper with distance. Light, bulky produce fills a vehicle's volume before its weight limit (it cubes out), so carriers also price by density.
Fleet costs divide into fixed costs, incurred whether the vehicle moves or not (depreciation, interest on capital, insurance, permits and taxes, and often the driver's wage), and variable costs, which rise with distance (fuel, tyres, maintenance, tolls and trip allowances). Cost per kilometre falls as annual use rises, so utilisation is the key lever.
Key measures are vehicle utilisation, load factor, empty running (often the return leg), cost per tonne-kilometre, on-time delivery and damage. A 16-tonne truck on a 400-kilometre trip costing 32,000 in currency units costs 5 per tonne-kilometre if full, but 6.67 if it carries only 12 tonnes, which is why load planning and finding backhauls matter. Seasonal fleets, such as cane haulage for a sugar mill, are usually hired.
The own-or-hire decision compares the fixed cost of owning with the premium paid for hiring. If owning a truck costs 600,000 a year in fixed costs plus 18 per kilometre in variable costs, and a hired truck costs 30 per kilometre, the break-even is 600,000 / (30 - 18) = 50,000 kilometres a year; above that, owning is cheaper, provided the owned truck can actually be kept busy. Control of timing and hygiene, reliable hauliers and capital also matter.
The vehicle routing problem seeks least-distance routes within vehicle capacities and time windows, often by the Clarke-Wright savings heuristic. The savings from combining two customers i and j on one route instead of serving each by a separate trip from the depot is the distance depot-to-i plus depot-to-j minus i-to-j. Driving-time and axle-load rules differ between countries.
The grain chain usually runs from farm storage or field to a country elevator, procurement centre or primary market; then to a terminal or port elevator, a mill, a feed plant or a public buffer store; and then by rail, barge or ship to processors or export. Large exporters move grain in bulk by truck, unit train or barge and bulk vessel. In many developing countries grain still moves largely in bags (commonly of 50 kilograms or thereabouts), labour-intensive but flexible and cheap to set up.
Bulk handling cuts handling cost per tonne and losses from spillage and bag damage, but needs large investment and it mixes lots, which works against identity preservation of specialty grain. Drying to a safe moisture, which depends on grain, temperature and storage period, and aeration prevent mould and insects; hermetic bags control insects by limiting oxygen and suit smallholders.
Seasonality drives grain logistics: most of the year's crop reaches market within a few weeks of harvest, so transport and storage capacity are stretched at harvest and idle later, and freight rates rise at harvest. Public procurement and buffer-stock systems add large flows; India's public wheat and rice stocks use covered warehouses, silos and, when short of space, cover-and-plinth (CAP) storage in the open. Port throughput and the freight market govern export flows, and closure of one river, canal or strait shifts grain flows worldwide.
A warehouse does more than store. Its functions are holding seasonal stock between harvest and use; consolidation, combining small inbound shipments into larger outbound loads; break-bulk, splitting large inbound loads into small outbound deliveries; mixing, assembling orders of many products from many suppliers; cross-docking, moving goods through with minimal storage; value-added services such as cleaning, grading, sorting, fumigation, ripening, packing and labelling; and, for agricultural commodities, serving as the basis for warehouse receipt finance and for delivery against commodity futures contracts.
Warehouses are classified by ownership: private (owned and run by the firm that owns the goods), public (run by a business or public body that stores goods for any depositor for a fee) and contract (dedicated space and services run by a logistics provider under a longer-term contract). By design there are distribution centres (throughput over storage), bonded warehouses (imports before duty), cold stores, silos and rural godowns. In India, the Central Warehousing Corporation and State Warehousing Corporations provide public warehousing, a central food agency holds public grain stocks, and a regulatory authority set up under the Warehousing (Development and Regulation) Act, 2007 registers warehouses that issue negotiable warehouse receipts.
Operations run receiving (checking, weighing, sampling), put-away, storage, order picking (usually the most labour-intensive step), packing and dispatch. Layout minimises travel and handling, keeps flow one-way or U-shaped, slots fast movers near dispatch, uses height where possible and separates incompatible goods such as chemicals from food.
Stock rotation rules are first-in first-out (FIFO), the default, and first-expired first-out (FEFO), which is the rule for perishables and products with a best-before or use-by date; last-in first-out (LIFO) is used physically only where rotation does not matter, as in some bulk piles. Capacity is measured in tonnes, bags, cubic metres or pallet positions, and a 1,000-tonne godown for 50-kilogram bags holds 20,000 bags at full weight capacity; in practice stacking plans, aisles and inspection space reduce the usable figure. Measures include space utilisation, throughput, inventory accuracy, picking accuracy, storage losses and cost per tonne-month.
Scientific storage uses clean, dry, rodent-proof buildings, stacking on dunnage away from walls, inspection, fumigation by trained operators under national rules, and moisture monitoring, against losses to insects, moulds and mycotoxins, rodents and theft.
A warehouse receipt is a document issued by a warehouse operator acknowledging that specific goods of a stated quantity and quality have been deposited. If negotiable, it can be transferred by endorsement to sell the stored goods without moving them, or pledged to a bank as collateral for a loan, so farmers and traders can obtain credit against stored produce. Warehouse receipt systems need trustworthy, licensed and inspected warehouses, reliable grading, insurance, and a legal framework that gives the receipt holder clear title. They are promoted widely to reduce distress sales at harvest.
First-party logistics is a firm moving its own goods; second-party logistics is hiring a carrier or warehouse for a single service. A third-party logistics provider (3PL) is an external firm that performs a set of logistics activities for a client under a contract, often integrating transport, warehousing, inventory management, cross-docking, packaging and information services. 3PLs may be asset-based (owning trucks and warehouses) or non-asset-based (managing capacity they buy from others). A fourth-party logistics provider (4PL), sometimes called a lead logistics provider, is an integrator that manages the client's whole logistics network, including several 3PLs, and brings technology, planning and management rather than physical assets; the term was coined by consultants in the 1990s.
Firms outsource to focus on core competence, turn fixed costs into variable ones, share a provider's scale economies (full loads, backhauls, shared warehouses), gain specialist capability such as cold chain or customs, and handle seasonal peaks. Risks are loss of control over service, dependence, information leakage, hidden costs and lost know-how. It is a make-or-buy decision governed by a service-level agreement with measurable KPIs (on-time delivery, inventory accuracy, temperature compliance), incentives and penalties. Agri 3PLs offer cold chain, collateral management for lenders, grain handling and export logistics.
Reverse logistics is the process of planning, implementing and controlling the flow of goods, materials and information from the point of consumption back towards the point of origin to recapture value or ensure proper disposal; Rogers and Tibben-Lembke's definition is the one most textbooks quote. In agri-food it covers returns of unsold or damaged product (common under sale-or-return terms for bread and dairy); recalls; returnable packaging such as crates, pallets and milk cans; rinsed pesticide containers taken back under stewardship schemes; by-products such as whey, bran, oilcake and bagasse; and food loss and waste.
The food waste hierarchy used by many governments ranks options from most to least preferred: prevention; redistribution of surplus food to people; use as animal feed where the law allows (rules on feeding food waste to livestock differ strictly between countries because of disease risk); recycling through composting or anaerobic digestion; energy recovery; and disposal. Reverse flows are less predictable, arrive in small mixed lots, need sorting before disposition and often have low value.
A recall is the reverse flow that matters most for food safety. Its success depends on traceability: knowing which lots went to which customers so that only affected product is withdrawn, quickly. Closed-loop chains, such as pooled reusable crates for fresh produce, integrate forward and reverse flows by design.
A cold chain is an unbroken series of refrigerated storage and distribution activities that keeps a product within a specified temperature range from harvest or slaughter to consumption, for produce, flowers, dairy, meat, fish, frozen foods and also vaccines.
Harvested produce is alive: it respires, giving off heat, loses water by transpiration, and ripens and senesces. The rate of respiration and most deterioration rises steeply with temperature: a common textbook rule is that the rate roughly doubles or triples for each 10 degrees Celsius rise within the range of normal handling (the temperature quotient, Q10). Cooling is therefore the most effective way to slow deterioration, and time spent warm is shelf life lost for good. Most fresh produce is kept at high relative humidity to limit water loss; onions and dried products need drier air.
Two biological cautions make cold chains more than simply cold. First, chilling injury: many tropical and subtropical products, such as bananas, mangoes, papayas, many cucurbits, and tomatoes, are damaged by temperatures above freezing but below a crop-specific threshold. Safe temperatures are crop-specific and are given in published storage guides. Second, ethylene: climacteric fruits such as apples, bananas, mangoes, avocados and tomatoes produce ethylene as they ripen, and ethylene accelerates ripening, yellowing and senescence in sensitive products such as leafy greens, cucumbers, carrots and flowers. Compatibility groupings keep such products, and products with different temperature needs, apart in mixed loads.
Harvest in the cool of the day and field shading reduce field heat. Pre-cooling removes field heat quickly after harvest and is the most important and most often missing link in developing countries. Methods are room cooling (slow), forced-air cooling (cold air pulled through vented packs), hydrocooling (chilled water, for produce that tolerates wetting), vacuum cooling (suited to leafy vegetables), and package or top icing. Milk is chilled in bulk coolers at farm or collection centre, to temperatures and time limits set by national standards.
Then come cold storage (short-term rooms, long-term stores for potatoes and apples, frozen stores), controlled-atmosphere storage (low oxygen and raised carbon dioxide at low temperature, which greatly extends the storage life of apples and some other fruits), refrigerated transport (reefer trucks, rail wagons and containers), ripening chambers where ethylene is applied under control to ripen bananas and mangoes evenly, retail display and home storage. Frozen foods are commonly held at minus 18 degrees Celsius or colder; chilled foods are held within ranges set by national food regulations and product specifications.
A cold chain is only as good as its weakest link: a few hours on a hot loading dock can undo days of good storage. Data loggers and time-temperature indicators (labels that change colour with cumulative exposure) monitor it. Some importers require cold treatment (fruit held cold for a set number of days) as a quarantine measure against fruit flies, logged in the reefer.
The economics of the cold chain turn on the value of shelf life. Cold chain investment is justified when the reduction in loss and the price premium from better quality and longer selling seasons exceed the capital and energy costs. A simple example: if 20 tonnes of a fruit selling at 30 per kilogram lose 25 percent in an ambient chain and 8 percent in a cold chain, the cold chain saves 3.4 tonnes, worth 102,000, which must be set against the cost of refrigeration, packaging and management for that lot. Utilisation matters: a cold store full for two months a year rarely pays.
In many developing countries the gaps are at the first mile (pack-houses and pre-cooling near farms) and in refrigerated transport, rather than in bulk cold storage, which has often been built for a few crops such as potatoes. Unreliable power, small lots and markets that do not pay for quality hold back investment; responses include public support, solar and evaporative coolers, and pay-per-use cold rooms run by FPOs. Loss figures for perishables vary widely by product, country and study method, so any single percentage should be treated with care; FAO's work distinguishes food loss (in production, post-harvest and processing) from food waste (at retail and consumption).
Information is the driver that lets a chain hold less stock, run fuller vehicles and respond faster. Enterprise resource planning (ERP) puts orders, procurement, production, inventory and finance on one database; warehouse management systems (WMS) track stock by location and lot; transportation management systems (TMS) plan loads and routes, select carriers and track shipments; planning software optimises networks and inventory.
Electronic data interchange (EDI) is the computer-to-computer exchange of standard documents such as purchase orders, advance shipping notices and invoices. Automatic identification uses barcodes and radio-frequency identification (RFID). GS1 standards identify a product (GTIN), a logistic unit such as a pallet (SSCC) and a location (GLN). RFID tags can be read without line of sight and in bulk, but cost more than barcodes and read less reliably through water and metal, which matters for food. Satellite tracking and internet-of-things sensors report vehicle position, reefer temperature and grain-bin conditions in real time.
Agri-specific systems include procurement apps recording farmer deliveries, quality and payments, mobile payments, electronic markets and satellite data for supply forecasts. Blockchain, a shared, append-only ledger, has been piloted for food traceability and trade documents; it makes records hard to alter after entry, but it cannot verify that what was entered was true, so it depends on the same data capture and audits as any other system. Evidence of its advantage over conventional databases in food chains is still limited.
Traceability, in the Codex Alimentarius definition, is the ability to follow the movement of a food through specified stage or stages of production, processing and distribution. Tracking (tracing forward) follows a lot downstream to find where it went (for recalls); tracing (tracing back) follows it upstream to find its origin (for outbreaks and complaints). Internal traceability links inputs to outputs inside one firm (which raw lots went into which finished batch); chain traceability links firms.
The basic legal model, required by the EU General Food Law, Regulation (EC) No 178/2002, is one step back and one step forward: each business identifies its immediate suppliers and business customers. The United States Food Safety Modernization Act led to a food traceability rule requiring businesses handling listed high-risk foods to record key data elements at critical tracking events (harvesting, cooling, packing, shipping, receiving, transformation) with a traceability lot code; its compliance date has been revised, so current dates should be checked. ISO 22005 sets principles for traceability systems in feed and food chains. Private certification schemes and some export markets require farm-level records, for example for residues or proof that a commodity was not grown on recently deforested land.
The unit of traceability is the lot or batch: a quantity produced under practically the same conditions. Smaller lots mean smaller recalls but costlier segregation. Bulk commodities such as grain and milk are mixed at every stage, so traceability is by batch and by time window rather than by individual farm, unless the chain is run as identity-preserved, where lots are kept separate at a cost. Livestock systems use individual animal identification and movement records, with designs that differ by country.
A traceability design starts from its purpose (recall, quality or sustainability claims, market access, disease control), sets the unit and depth, identifies the critical tracking events where lots are created, split, mixed or moved, specifies the key data elements to capture at each event (who, what, where, when, lot code, quantity), and chooses the carriers of identity (labels, barcodes, RFID, documents) and the data system. It must then be tested with mock recalls, in which a firm traces a chosen lot forward and back against the clock to show it can find all affected product.
For smallholder chains, the hardest step is the first mile: recording which farmer delivered which quantity and quality, and keeping lots from different farmers separate or at least recorded when they are pooled. Group certification through FPOs, collection-centre records, simple mobile applications and labelled crates make this feasible. Systems last when they also pay back through faster payment, fewer disputes and better forecasts, not only compliance.
Use this subject to answer the operations part of agribusiness entrance and course exams and to read a firm's logistics numbers. Practise until you can compute an EOQ, a reorder point with safety stock, a newsvendor order and a moisture shrink without a formula sheet, and until you can explain the bullwhip effect, cross-docking, pool distribution and one-step-back-one-step-forward traceability in two sentences each. On a summer internship at a processor, input company or retailer, map the physical chain from farm to shelf for one product, mark every handling and storage point, and estimate where time, temperature and money are lost; that map is the start of any useful project report.
A procurement, logistics or supply-chain manager five years in sets reorder points and stocking norms for depots and dealers before each season, negotiates freight and 3PL contracts with measurable service levels, decides own-versus-hire for vehicles, plans harvest-time procurement centres and storage, and runs mock recalls. The daily decisions are how much to stock where, how to fill vehicles and find backhauls, which suppliers or farmer groups to contract with and on what quality and price terms, and how to keep cold-chain temperatures in range. The tools are the forecasting and inventory formulas of this manual, a spreadsheet or ERP, a transport management system and simple dashboards of fill rate, inventory days, cost per tonne-kilometre and loss.
A senior executive with twenty-five years of experience designs the network: how many plants, collection centres, warehouses and cold stores, where, owned or outsourced, and how much buffer capacity to hold against droughts, disease outbreaks and route closures. They choose between spot, contract and own supply for strategic raw materials, set the balance between an efficient chain for commodities and a responsive one for fresh and branded lines, and decide investments whose payback depends on utilisation, such as silos, pack-houses and reefer fleets. They also answer to regulators and buyers on traceability, food safety recalls and sustainability claims, and they build the data-sharing relationships with retailers and farmer organisations that reduce the bullwhip effect.
A researcher with long field experience works on questions this manual only outlines: measuring post-harvest loss consistently, modelling perishable inventory and cold-chain investment under uncertain power and prices, designing contracts that reduce side-selling and opportunistic rejection, estimating the welfare effects of modern retail and electronic markets on smallholders, testing whether traceability technologies such as blockchain add value over conventional records, and optimising multi-echelon networks with stochastic supply. Methods include stochastic and robust optimisation, simulation, econometric impact evaluation and field experiments. Such a researcher also advises governments on storage policy, warehouse receipt law and cold-chain subsidies, where the evidence base is thin and the stakes for food security are high.
This manual does not yet cover multi-echelon inventory optimisation, the (s, S) and base-stock policies in formal detail, the transportation and assignment problems of linear programming worked step by step, full vehicle-routing algorithms, or supply-chain finance beyond warehouse receipts. It gives no crop-by-crop storage temperatures, humidity settings or storage lives, because these vary by product, cultivar, maturity and source; use a published postharvest storage guide. It gives no freight rates, storage tariffs, logistics cost shares or post-harvest loss percentages, because published figures vary widely by country, year and method. Country laws on contract farming, warehousing, food traceability and driving hours are summarised only in outline and must be checked against current national texts; the compliance timetable of the US food traceability rule in particular has been revised. Sustainability metrics (carbon footprint of logistics, refrigerant rules) are not covered.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions. Every numerical answer was computed by script; figures are illustrative, not current prices or rates.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.