Losing on the clock
They supply fasteners, seals and fittings to contractors across the Northern Emirates. Twenty-seven thousand line items across three warehouses.
Quoting worked like this. A contractor emails a bill of quantities. An estimator opens the price list workbook, cross-references a stock report exported from the accounting system that morning, phones the Ajman warehouse to check whether the report is telling the truth, applies a customer discount tier held mostly in his head, and types a quote into Word.
Two days was normal. Three during Ramadan or when the estimator was on leave.
Their operations director had been told repeatedly by customers that the prices were fine and the response times were not. Two competitors were quoting inside a day.
What the spec sprint found first
We spent eleven days on specification. The most valuable thing it produced wasn't the design.
It was the discovery that their accounting data was in worse condition than anyone had admitted out loud: duplicated SKUs across warehouses under different codes, roughly 900 items with no reliable cost price, and a customer table where the same contractor appeared four times.
Building a quoting engine on top of that would have produced fast, confident, wrong quotes.
So the build decomposition opened with a data remediation milestone, priced and scheduled like any other. Found in week two of planning rather than week six of building, which is the difference between a line item and a crisis.
Delivered
AED 118,000 fixed. Nine milestones. Ten weeks including remediation.
Estimators now open a bill of quantities, get live stock across all three warehouses, discount tiers applied automatically by customer record, and a branded PDF out the other side.
| Measure | Before | After 6 months |
|---|---|---|
| Average quote turnaround | 2 days | 35 minutes |
| Quotes issued per month | 88 | 214 |
| Tender win rate | 19% | 27% |
| Estimator hours on quote prep, weekly | 31 | 9 |
| Stock discrepancies flagged at pick | ~40/month | 6/month |
Payback landed at about five months on additional gross profit alone, before counting the estimator time that went back into supplier negotiation.
The operations director, on the data milestone: "I'd have argued against paying for that if you'd asked me in month one. It turned out to be the only reason the rest of it works."