Procurement Insights Suite

Business Question

Procurement conversations tend to run on totals — total purchase amount, total pending. The real questions live one level down: which items and vendors drive the spend, where are orders stuck, what does unit cost actually do month to month, and how much inventory is sitting where? This suite was built to get to that root-level data.

Data Sources

Purchase orders, invoices, vendor master and inventory movement from the ERP system, refreshed on a reporting cadence and modelled into a consistent item / vendor / location structure.

KPIs

Purchase amount, consumption and inventory value; vendor counts and top-10 concentration; average order quantity and unit price; pending orders in both quantity and value; per-vendor fulfilment rates and days-since-last-purchase; and days-of-inventory with and without pending orders.

Dashboard Design

The suite opens on an index page and fans out into three analysis pages — Procurement, Vendors, and Items — each drillable by item group, vendor and financial year.

The views below are recreations with fully synthetic data — all vendor names, sites and values are anonymized.

Key Responsibilities

• Framing the analysis with procurement stakeholders — what "root level" actually needed to mean
• Designing and building all four pages of the Power BI suite
• Modelling item, vendor and location hierarchies consistently across pages
• Building fulfilment-rate and days-of-inventory measures in DAX
• Iterating the views based on review-meeting usage

Technologies I Personally Used

Insights

The suite made procurement concentration explicit — a handful of items and vendors accounting for most spend — and separated price effects from volume effects by tracking unit cost as its own trend rather than letting it hide inside purchase totals. Pending-order value per vendor turned a vague "follow up with vendors" into a ranked list.

Business Decisions

Reviews shifted from totals to specifics: which vendors to chase on outstanding orders, which items justified renegotiation given unit-cost movement, and where inventory was accumulating relative to consumption.

Business Impact

Business impact is currently being quantified and will be updated as validated metrics become available. Qualitatively, procurement review now runs on one suite instead of assembled spreadsheets.

Lessons Learned

• "Root-level" reporting is a modelling problem before it's a visual one — the item/vendor/location hierarchy has to be right once, centrally
• Unit price deserves its own chart; buried in totals, cost movement goes unnoticed
• A ranked pending-by-vendor list gets acted on; a pending total does not

Reflection

What I learned: procurement data rewards patience — the value is in consistent definitions across four pages, not in any single visualization.

What I would improve today: add vendor lead-time analysis (order date to receipt date) as a first-class measure alongside fulfilment rate.

Future enhancements: price-variance alerts on key items, vendor scorecards combining fulfilment, lead time and outstanding behaviour, and integration with inventory reorder planning.