Case: Warehouse Network Redesign - Interview Framework, Metrics And Mini Case

A institution tin spell from “next-day transportation successful astir cities” to “orders delayed, trucks half-empty, inventory stuck successful the incorrect place” without changing a azygous product. The culprit is often not the storage itself, but the web - wherever inventory sits, really it flows, and which customer committedness each node is meant to serve.

Warehouse web redesign intends deciding the number, location, domiciled and capacity of warehouses to meet work targets astatine the lowest full cost.
Do not optimise only carrier cost. The existent trade-off is transport + warehousing + inventory + work + modulation risk.
A beardown lawsuit reply starts pinch request density and work promise, past designs nodes, flows, inventory rules and migration.
The classical options are centralised, decentralised and hybrid networks. Most existent companies usage a hybrid.
Use metrics for illustration costs per order, OTIF, mean transportation distance, capacity utilisation, inventory turns and bid rhythm time.
A redesign is not “open much warehouses.” More nodes trim last-mile region but summation fixed cost, information banal and coordination complexity.
The winning reply is simply a segmented network: fast-moving SKUs adjacent to dense demand, slow movers centralised, and exceptions handled deliberately.

The Big Picture

A storage web is the beingness skeleton of service. Redesigning it is astir matching where request is, how accelerated customers expect delivery, and what it costs to clasp and move inventory.

A storage redesign moves the institution from 1 generic web to differentiated flows for different customers and SKUs.A storage redesign moves the institution from 1 generic web to differentiated flows for different customers and SKUs.BeforeOne-size networkAfterSegmented by serviceA storage redesign moves the institution from 1 generic web to differentiated flows for different customers and SKUs.

Core Explanation: How to Solve a Warehouse Network Redesign Case

Think of the lawsuit arsenic a total-system redesign. A storage is only 1 node. The existent strategy includes suppliers, factories, ports, mother warehouses, location distribution centres, acheronian stores, stores, transportation partners and customers.

The cleanable consulting measurement to lick it is to move from request to design, not from creation to demand. In different words, do not commencement pinch “we request 3 warehouses.” Start pinch “who needs what, where, and really fast?”

A beardown warehouse-network reply follows this series alternatively of jumping consecutive to storage count.A beardown warehouse-network reply follows this series alternatively of jumping consecutive to storage count.MapDemandWhereorders…SetPromiseSpeed andreliabilityDesignNodesNumberand rolesModelCostTotalsystem…MigrateSafelyPilot thenscaleA beardown warehouse-network reply follows this series alternatively of jumping consecutive to storage count.

Step 1: Map request by geography, customer and SKU

Split request into useful clusters: city, pin code, channel, customer type and SKU velocity. A redesign for B2B spare parts is not the aforesaid arsenic a redesign for grocery, manner aliases appliances.

Look for 3 patterns:

Demand density: Are orders concentrated successful a fewer metros aliases dispersed crossed galore towns?
SKU velocity: Which products are fast-moving, slow-moving, bulky, vulnerable aliases high-value?
Service sensitivity: Which customers genuinely request same-day aliases next-day delivery, and which tin wait?

Step 2: Define the work committedness earlier the storage count

Service committedness is the target experience: same-day, next-day, two-day, play replenishment, capable complaint aliases emergency availability. The work target decides whether inventory must beryllium adjacent to the customer aliases tin beryllium pooled centrally.

If the committedness is loose, centralisation usually wins because inventory pooling reduces duplication. If the committedness is tight, proximity matters, but costs and information banal rise.

Segmenting request prevents the communal correction of forcing each bid done the aforesaid storage logic.Segmenting request prevents the communal correction of forcing each bid done the aforesaid storage logic.Premium localDense and urgentRegional FCDense but patientException modelSparse but urgentCentral poolSparse and patientDemand densityService urgencySegmenting request prevents the communal correction of forcing each bid done the aforesaid storage logic.

Step 3: Choose the web archetype

Most redesign cases inquire you to comparison 3 archetypes. The interviewer is checking whether you understand the trade-off, not whether you memorise jargon.

Step 4: Model full cost, not conscionable carrier cost

A storage redesign must see each awesome costs buckets. If you trim outbound freight but double inventory and rent, the redesign has failed.

Step 5: Test feasibility and modulation risk

A beautiful web connected Excel tin neglect successful reality. Check whether the institution has the land, labour, carrier partners, storage guidance system, supplier subject and change-management capacity to execute the redesign.

This is wherever a consulting reply becomes practical. If you want the broader discourse of really specified activity moves from test to implementation, revise the engagement lifecycle from kickoff to handover.

Worked Example: First-Pass Warehouse Location

Suppose a institution serves 3 request zones. A elemental centre-of-gravity calculation gives a first-pass location, weighted by bid volume. It is not the last answer, but it is simply a useful starting constituent earlier checking roads, rent, work promises and constraints.

Weighted X = (20 x 200 + 10 x 300 + 90 x 500) / 1,000 = 52.

Weighted Y = (80 x 200 + 20 x 300 + 30 x 500) / 1,000 = 37.

So the first-pass location is adjacent (52, 37). But an interview-ready reply adds the caveat: “I would now overlay roadworthy connectivity, real-estate cost, labour availability, supplier routes, work promises and risk.”

Asian Paints is often discussed arsenic a supply-chain benchmark because its distribution spot is not conscionable storage placement. The superior driver is simply a tightly managed replenishment and dealer-service system, supported by request visibility, manufacturing responsiveness and disciplined logistics execution. The truthful what: storage web creation wins erstwhile it is connected to work committedness and readying systems, not treated arsenic a standalone real-estate decision.

Definitions You Can Say successful One Breath

Warehouse web design: The prime of storage number, location, domiciled and capacity to meet work astatine minimum full cost.
Cost to serve: The afloat costs of fulfilling a customer aliases segment, including storage, handling, transport, inventory and work exceptions.
Service level: The probability aliases proportionality of request fulfilled wrong the promised time, amount and condition.
Inventory pooling: Holding banal centrally truthful request variability is shared crossed locations, reducing duplicated information stock.
Safety stock: Extra inventory held to protect against request uncertainty, proviso delays aliases forecast error.

BigBasket: Redesigning the Grocery Network for Speed

BigBasket shows really market fulfilment moved from scheduled transportation logic toward a much proximity-led web arsenic speedy commerce reshaped customer expectations successful India.

Grocery velocity is won by placing the correct inventory person to dense section demand.

Situation. Online market has a difficult supply-chain problem: request is frequent, baskets are mixed, freshness matters, and customers progressively expect speed. A purely centralised storage tin excavation inventory efficiently, but it struggles erstwhile customers want accelerated transportation crossed dense metropolis pockets.

The move. BigBasket’s quick-commerce play required a much section fulfilment logic: smaller neighbourhood nodes for fast-moving items, larger upstream accommodation for replenishment, and sharper assortment decisions by micro-market. The superior driver was proximity to demand. Supporting drivers included amended request forecasting, curated section assortment, rider routing, replenishment subject and cold-chain handling wherever needed.

The lesson. The redesign was not simply “add much warehouses.” It was a network-role change. Large accommodation remained useful for pooling and replenishment, while section nodes served urgent, high-frequency demand. That is the nonstop penetration interviewers want: redesign the web by segment, not by instinct.

BigBasket-style speedy commerce depends connected aggregate operating levers, not section warehouses alone.BigBasket-style speedy commerce depends connected aggregate operating levers, not section warehouses alone.Local NodesFast movers nearbyForecastingPin-code demandMother HubsBulk replenishmentRoutingShort transportation loopsCustomer PromiseBigBasket-style speedy commerce depends connected aggregate operating levers, not section warehouses alone.

How AI Changes Warehouse Network Redesign

AI changes this taxable by making the redesign much dynamic. Earlier, companies redesigned storage networks each fewer years. Now, request shifts, transportation promises and inventory placement tin beryllium simulated and adjusted much frequently.

Granular request forecasting: Machine learning tin forecast request by pin code, SKU, time of week, upwind shape aliases section event. This improves which SKUs should beryllium successful section nodes versus cardinal pools.
Scenario simulation: Optimisation models tin comparison “2 warehouses versus 5 warehouses,” different work promises, rent assumptions, substance costs sensitivity and stock-out consequence earlier beingness changes are made.
Smarter inventory placement: AI tin urge wherever to position inventory based connected probability of sale, lead time, capacity constraints and substitution patterns.

Use ChatGPT aliases Claude to practise: paste a short institution description, request clusters, work targets and costs buckets, past inquire it to make 3 web options pinch trade-offs, assumptions and risks. Then situation the output by asking: “Which hidden costs aliases operational constraint person we missed?”

Interview Relevance

“An e-commerce institution has rising transportation costs and inconsistent work crossed India. It presently uses 1 nationalist storage and 3 location warehouses. How would you redesign its storage network?”

Say this statement successful the interview: “I will optimise for full landed work cost, not storage costs alone.” It signals maturity immediately.

Common Mistake

The biggest correction is treating storage redesign arsenic a elemental “how galore warehouses?” problem. That misses inventory duplication, work segmentation, inbound flows, exertion readiness and modulation risk. The one-line fix: start pinch request and work promise, past creation the web that serves each conception astatine the lowest full strategy cost.

What to Revise Next

Next, revise Case: Reducing Cost to Serve Without Hurting Service. Warehouse redesign is 1 powerful lever; the adjacent lawsuit teaches you really to trim fulfilment costs crossed customers, channels and products without damaging the committedness that customers really value.