The site store is where 60% of project cost passes through. GRN discipline, task-linked issue slips, min-max levels and quarterly stock-takes that hold.

Materials are typically 55 to 65 percent of a civil project's cost, and every rupee of it physically passes through one place: the site store and godown. Yet on most sites the store is the least managed function on the project — a container and a tin shed run by the most junior person on the payroll, on registers that get filled in before audits rather than during work. The gap between a loose store and a tight one is worth 3 to 5 percent of material cost — on ₹4 crore of annual material spend, ₹12–20 lakh a year, silently. Tightening it does not need software first; it needs four disciplines, which software then makes permanent.

First, Treat the Store Like the Bank Vault It Is

Basics that cost nothing and change everything: a defined custody boundary (nothing enters or leaves except through the store keeper — material dumped at the gate by a supplier is not "received" until the store says so); segregated storage that protects value — cement on dunnage, off the floor and covered, issued strictly first-in-first-out because it is a perishable with a shelf life; steel on racks by diameter, tagged by heat number for traceability; diesel in a locked dispensing arrangement with a log; and cameras on the gate and the fuel point, less to catch anyone than to make the honest path the easy one. And one organisational rule: the store keeper reports to the project manager, not to the engineer whose consumption he is recording — measurement and execution should never share a boss.

Discipline 1: GRN — Nothing Exists Until It Is Received Properly

The Goods Receipt Note is the store's front door, and every leak walks through it when it is left open. The rules:

Every receipt against a purchase order. Material arriving without a PO is a phone-call purchase — the pricing discipline died before the truck arrived.
Count and inspect before signing, not after. Cement in bags counted and weighed on sample; steel weighed on a weighbridge, not accepted on the challan's word — the difference between challan weight and actual weight on reinforcement, compounded over a project, funds someone's second business; aggregate volumes checked against truck body measurement.
Record shortages and rejections on the GRN itself, photographed, and communicated to the supplier the same day — a shortage recorded a week later is a favour, not a claim.
GRN before invoice booking, always. Accounts pays against the three-way match — PO, GRN, invoice. The moment an invoice can be booked without a GRN, your payables are decided by suppliers instead of by your store.

The test of GRN discipline is timing: a GRN made at the tailgate while the truck waits is a record; a GRN made in the evening from challans is fiction with a signature.

Discipline 2: Issue Slips — Consumption Gets a Name and a Purpose

Receiving well means little if issues are loose. "Give Ramesh 20 bags" is how theoretical-versus-actual reconciliation dies. Every issue needs three attributes: who (the gang or subcontractor drawing it), what for (the task or BOQ item it will be consumed against), and authorisation (an indent approved by the engineer, with value limits above which approval escalates). Task-linked issues are what make consumption analysis possible at all — when 50 bags are issued against the grid-C raft, the system can compare actual consumption with the theoretical norm for that concrete volume within days, and an 8.6-bags-per-cum site against a 7.4 norm gets caught the same week instead of at project close. Three supporting rules: returnable gate passes for anything leaving site — shuttering plates to another site, a motor for rewinding — logged with expected return dates and reconciled weekly, because "returnable" without reconciliation is a euphemism for gone; returns to store recorded as scrupulously as issues, or surplus steel at slab-end becomes invisible inventory rusting behind the labour colony; and daily closing stock for A-class items — cement, steel, diesel — recorded every evening, making the store's arithmetic continuous instead of episodic.

Discipline 3: Minimum Stock Levels — Neither Stockout Nor Graveyard

A store fails in two opposite directions: the stockout that idles a 40-worker gang at ₹700 a head while everyone blames purchase, and the over-stocked godown quietly burying working capital in materials bought "to be safe" — some of which will be damaged, expired or stolen before use. The way out is classification and levels. Run an ABC analysis: A-class items (cement, steel, diesel, structural fittings) are few but carry most of the value — they get daily stock visibility, defined minimum and maximum levels, and reorder points set from genuine lead times (a vendor who delivers in five days needs a different trigger than a special fitting on three weeks). B and C class get simpler min-max rules and periodic review. Two habits keep levels honest: recompute them when the programme changes — a slab-a-week tower in structure needs different cement cover than the same tower in finishing — and review non-moving stock monthly, because every item with no issue in 90 days is a decision waiting: return it, transfer it to a site that needs it, or admit the purchase was a mistake while the vendor still takes returns.

Discipline 4: The Quarterly Stock-Take — Trust, Verified

Perpetual records drift from physical reality; the stock-take is where they reconcile. Quarterly for the full store, monthly for A-class, and occasionally unannounced — the schedule everyone knows produces the count everyone prepared. Run it properly: blind counts (the counting team records what it finds without seeing book balances first), counters who are not the custodians, and every variance beyond tolerance investigated to a cause — receipt not recorded, issue not recorded, wastage, theft — rather than "adjusted". Write off what is genuinely lost through a named approval, because a register that silently absorbs shrinkage is a register that invites it. Then close the loop: feed reconciled consumption back into project costing and into your estimation norms, so the next tender is priced on what sites actually consume, not on what the handbook hoped.

Do Not Forget the Unglamorous Ledgers

Two registers separate a professional store from a tidy one. The tools and plant register: vibrators, cutting machines, drills, pumps and survey instruments issued to gangs against signature, with a monthly physical verification — small assets leak faster than materials because nobody prices them until twenty are missing. And the scrap ledger: steel offcuts, empty cement bags, used shuttering and drums all have market value and a well-established habit of monetising themselves informally; weigh scrap out through the gate against authorised sale notes, and the proceeds land in the project's account instead of the neighbourhood's. Diesel deserves its own line: every litre dispensed logged against a machine and its hour-meter reading, because fuel is the A-class item that walks fastest of all.

Making It Permanent

Paper can hold these disciplines for a quarter; software holds them for the life of the company. A digital store — GRNs against POs at the tailgate, indent-approve-issue flows on a phone, task-linked consumption, live min-max alerts and stock-take reconciliation — is exactly what inventory & supply runs, with theoretical-versus-actual computed daily instead of discovered at audit. See how contractors structure the wider material cycle, purchase to consumption, in our material management solution, and put your own material spend into the ROI calculator to price what 3 percent leakage is costing you today.

The store keeper with a system is the cheapest margin protection a contractor can hire.

Book a ProjectSathi demo and see GRNs, issue slips and stock levels running live on a site store.