How a Procurement Team Uncovered $2.8M in Supplier Risk - in 4 Minutes
The Situation
A mid-size supply chain consultancy was preparing a quarterly procurement review for a manufacturing client. The client had 11,400 purchase orders across three files - one per quarter - covering raw materials, electronics, chemicals, packaging, and MRO supplies.
Their usual process: export from ERP, clean the data in Excel, build a pivot table, manually create supplier rankings, write a summary. Two analysts. Half a day. Every quarter.
The files had inconsistent column names across quarters - Q1 used “Supplier_Name”, Q2 used “vendor”, Q3 used “SUPPLIER”. No standard schema. Every BI tool they'd tried either failed to load the files or required manual column mapping before it would do anything. They dropped all three files into Xnorly.
What Xnorly Found in 4 Minutes
Xnorly detected the schema across all three files automatically - no mapping, no templates - joined them into a single dataset, and generated the full analysis.
"Your top 5 suppliers account for 86.7% of total spend. $2,860,000 is at risk from single-source dependency."
The Pareto analysis revealed that NovaTech Components alone represented 30.8% of total procurement spend - $1,018,400 across the period. The top 3 suppliers combined controlled 68% of spend. No alternative sourcing was visible in the data for any of these categories.
Xnorly flagged this as high-priority with a specific dollar figure attached: if NovaTech experienced a supply disruption, $1M+ in procurement would be exposed with no fallback supplier in the dataset.
"Supplier A (NovaTech) is delivering an average of 6.2 days late. This has worsened by 2.1 days quarter-over-quarter."
Xnorly cross-referenced order dates against delivery dates across all three quarters and detected a statistically significant deterioration in NovaTech's delivery performance - coinciding with a 22% increase in order volume in Q3.
The system also flagged that Starline Metals Ltd, the second-largest supplier, had perfect on-time delivery across all 847 orders - making it a candidate for volume reallocation.
"Class A items (top 20% by value) account for 78% of spend but only 31% of order frequency - these are your strategic procurement targets."
Xnorly automatically classified all SKUs by revenue contribution (ABC) and demand variability (XYZ). The analysis identified 43 Class AX items - high value, stable demand - that were being ordered ad hoc rather than on scheduled contracts, creating unnecessary price volatility.
It also identified 218 Class CZ items - low value, unpredictable demand - consuming 14% of the procurement team's processing time for less than 2% of spend value.
"Q3 Electronics spend increased 34% with no corresponding change in production output data. Possible unplanned purchasing or inventory build."
Xnorly detected a significant spend spike in Electronics that didn't correlate with the demand patterns from Q1 and Q2. Without production output data to confirm, it flagged this as a potential unplanned purchase or buffer stock build that warranted investigation.
Results Summary
“We dropped three messy files with different column headers from three different quarters. Xnorly joined them automatically and gave us the supplier concentration risk in dollar terms within minutes. That single finding opened a $180K consulting engagement. We used to spend half a day on this review. Now it takes one coffee.”
- Anonymous, Supply Chain Analyst, Boutique Procurement Consultancy
The Data
- Files3 quarterly procurement exports (CSV)
- Rows11,400 purchase orders
- Columns detected14 per file (inconsistent naming - auto-resolved)
- Suppliers analyzed47
- CategoriesRaw materials, Electronics, Chemicals, Packaging, MRO
- PeriodQ1–Q3 2024
- Setup requiredNone. Files dropped in, analysis generated.
Want to run this analysis on your procurement data?
Drop your files at app.xnorly.com - no account setup, no column mapping, no waiting.
Try Xnorly freeThis case study uses anonymised data from a real analysis run through Xnorly. Company and supplier names have been changed. All financial figures reflect the actual output of the analysis.