Setting low-stock thresholds that actually prevent emergency purchases
Emergency cement runs cost 20–30% more than planned orders. Here's how to set stock thresholds that trigger on time, not after the pour stops.
An emergency cement order on a Friday afternoon costs 20–30% more than the same order placed Tuesday morning. The premium is real: rush logistics, last-minute supplier margin, the small bribe the driver needs to come on a holiday. And the reason the order became an emergency is almost never "bad luck." It is a threshold that was set wrong — or not set at all.
The four numbers every threshold needs
A threshold is not a single number. It is four.
Current stock. How much is on site right now. Measured, not guessed.
Daily consumption rate. How much the site burns per day when work is active. Calculated from the last 14 days of consumption, not from the first week (which is always slower than the rest).
Lead time. How many days from "place the order" to "truck at the gate." For a regular supplier in Tashkent — usually 2 days. For out-of-region — 4–7. For imported — 21–45. This number is the one that changes most between suppliers and most people never update it.
Safety buffer. How many extra days of stock you want above the minimum, because something always goes wrong. 2 days is tight. 5 days is comfortable. 10 days is hoarding — and hoarding is a different waste.
The threshold is then: daily consumption × (lead time + safety buffer). When current stock drops to that level, re-order.
Example from an actual site
Cement consumption = 3 tons/day during active pour. Lead time from the usual supplier = 2 days. Safety buffer = 3 days. Threshold = 3 × 5 = 15 tons. When the warehouse falls to 15 tons, the warehouse manager triggers a purchase order — not when it falls to zero.
Without that math, the warehouse manager "keeps an eye on it" and notices at 8 tons on a Friday, and the emergency premium begins.
Thresholds have to be material-by-material
A sandbag threshold is not the same as a cement threshold. Rebar lead time is not the same as nail lead time. Each material needs its own four numbers. It sounds like a lot of work, but 80% of a site's emergency purchases come from 5–10 materials — the ABC list — and those are the ones that deserve real thresholds. Everything else can use a rule of thumb.
Thresholds have to move with the project phase
A site in foundation phase burns cement at one rate. A site in finishing phase burns it at a fraction of that. Thresholds calibrated for one phase fail in another. Review thresholds every time the dominant work type shifts — foundation to structure, structure to enclosure, enclosure to finishing. Four reviews per project covers most of what you need.
Who owns the threshold
This is the part most Uzbek firms get wrong. The threshold is not set by the prorab (who thinks in days of work, not days of stock). It is not set by the supplier (who benefits from frequent rush orders). It is set by the warehouse manager, with the director signing off on the top 10 materials.
The prorab owns consumption; the warehouse manager owns stock levels; the director owns the buffer decision (tight vs. comfortable). Three roles, one threshold.
What triggers on threshold
Automated if possible: a purchase request is generated the moment current stock crosses the threshold, addressed to the approved supplier, with the next delivery date pre-calculated. Manual if you have no system: a daily 2-minute check at close of day. Either way, the threshold trigger is not "I'll think about it" — it is "the order is placed by end of business today."
What a good threshold system produces
Three visible changes after 60 days. First — emergency orders drop to near zero. Second — the director stops getting Friday-afternoon panic calls. Third — the warehouse stops hoarding, because the threshold gives permission to hold less safety stock once the system is trusted.
Thresholds are unglamorous. They are also the single operational change that pays for itself in the first month on any site that did not have them before.