Process automation case study using the ROI-FIRST approach
Purchasing Management for Construction Sites
Industry
Engineering & Construction
Date
Sep 18, 2026
Recommendation
✅ Automate
Time saved
From 39 min to 14 min per purchase request
ROI Automation, recommendation and costs
COP
currency $:
Level of Automation | % Automation | Cases x Month | Time x Case (hours) |
|---|---|---|---|
High - public requisition forms with mandatory specifications, routed approvals and returns with recorded reasons, automatic handoff to Treasury, supplier legal checklist and site-entry document portal with alerts, and real-time management dashboards. Supplier selection by the purchasing committee remains human | 64 | 990 | 0.64 |
Personas Involucradas | Nivel de Cargo Promedio | Average Monthly Salary | Real Company Cost with Benefits |
|---|---|---|---|
14 | Mixed: General Manager (Level 1, x1), Coordinator/Supervisor (Level 3, x3), Senior Analyst (Level 4, x2), Site Resident Engineer (Level 4, x8) | 3.639.474 | 5.580.413 |
Monthly Operational Cost | Annual Operational Cost | Estimated Monthly Savings |
|---|---|---|
21.205.569 | 254.466.828 | 13.508.396 |
Estimated Annual Savings | Year 1 Investment | Net Benefit Year 1 |
|---|---|---|
162.100.746 | 9.180.000 | 152.920.746 |
Cumulative Net Benefit 3 Years | FTE Equivalente Liberado |
|---|---|
458.762.238 | 3.8 |
Power in Numbers
1666
ROI Year 1 (%)
0.7
Payback (months)
17.7
Cost/Benefit ratio
✅ Automate
64% of this process is automatable, weighted by each role's actual workload. Purchasing consumes 3.8 FTE across 14 people and COP $254,466,828 per year. Automating it saves COP $162,100,746 annually against a COP $9,180,000 Year 1 investment, a 1,666% ROI with a 0.7-month payback. The savings come mainly from the purchasing lead, the approving coordinators and the site residents, who today chase status and clarify specifications outside any system. But the deeper issue is control. One person manages 40 to 50 requests a day, and if he is absent the process stops. Incomplete specifications and short deliveries force full restarts, including the three-supplier comparison, and up to five trucks have been turned away at sites for missing paperwork. Without recorded returns, management cannot tell whether the delay is the supplier, the site or purchasing. Automating here gives the same team the data to plan instead of react.
Signs of inefficiency detected
High manual follow-up: every return, supplier query or step back is handled by email, WhatsApp or phone and is not recorded in the purchasing ERP. Critical dependency on one person: a single purchasing lead manages 40 to 50 requests per day; if he is absent, the process stops. Incomplete specifications: requisitions arrive without model, size or exact description, forcing returns before the supplier comparison. Full restarts: when a supplier delivers 5 of 20 items, the purchase is cancelled and the three-supplier comparison starts over. Unknown real time: the two-day target exists on paper but is not measured by stage or owner. No status visibility: requesters cannot see where their request is or why it was returned. Site-entry failures: vehicle documents are sent 8 hours before delivery instead of the required 24, and up to five trucks have been turned away. Supplier compliance: legal documents are checked on paper, with no annual audit trail.