Spend analysis guide: building the spend cube

What "the spend cube" means

A spend cube is a three-dimensional view of expenditure: supplier, category, business unit / cost centre. Some teams add a fourth dimension (time, for run-rate and seasonality). A professional body's reference page on spend analysis gives the working definition and the standard uses (baselining, opportunity sizing, compliance tracking) ^[1].

The idea itself is not new: AberdeenGroup's 2004 paper Best Practices in Spending Analysis remains a useful primer on why enterprises struggle to see their own spend, and what a good process looks like.

The data you need

To create a view from your own transactions, use our guide and tool for building a procurement spend cube. It includes column mapping, filters and a Caventis Excel template.

A workable spend analysis usually needs 12-24 months of transactional data from three systems:

• AP / ERP: supplier name, invoice line, amount, GL account, cost centre, date.
• P-card / expense: transaction detail for below-PO spend.
• Supplier master: to consolidate duplicate supplier records (Acme Ltd vs Acme UK vs ACME LIMITED).

The single biggest data quality issue is supplier duplication: a small amount of fuzzy-matching investment early has an outsized effect on the accuracy of every downstream analysis. Cleaning and classifying the data is where most of the work actually lives (see this practical walkthrough on spend data classification).

A category taxonomy that will not fall over

Pick a taxonomy standard rather than inventing one. The most widely adopted global standard is UNSPSC (United Nations Standard Products and Services Code): a four-level hierarchical scheme (segment → family → class → commodity) maintained by GS1 US. For UK public-sector work, ProClass and the Government Commercial Function's Common Areas of Spend taxonomies are also worth mapping to.

A pragmatic UK taxonomy usually has 8-12 top-level categories, each split into 3-8 sub-categories. Deeper than that and classification effort exceeds the marginal analytical value.

What good spend analysis outputs look like

• Pareto by supplier: the top 20 suppliers typically cover 60-80% of addressable spend in a UK business: the renegotiation shortlist.
• Pareto by category: which categories concentrate spend, which are fragmented (a fragmentation index is a fair proxy for consolidation opportunity).
• Tail spend view: the count of suppliers under a threshold (e.g. <£25k p.a.) and the aggregate spend they carry.
• Off-contract spend: transactions that bypass a preferred supplier; this is a compliance leak.
• Payment-term ladder: weighted-average DPO by category, to expose working-capital opportunities alongside price.

From spend analysis to action

Spend analysis on its own does not save money: it just tells you where to look. The value comes from prioritising the top 10-20 opportunities, sizing each, assigning owners and running the sourcing/renegotiation. See our procurement cost reduction strategies for the standard sequence, and our category management guide for how to hold the savings. Top-quartile procurement functions return 2.4x ROI: savings worth more than twice the cost of running the function ^[4], and that gap is built on exactly this kind of prioritisation discipline, not on better negotiating alone.

Spend analytics as a continuous management routine

The Digital World Class® Procurement benchmark consistently shows that top-performing procurement organisations differ from typical peers primarily in their use of digital tooling, including continuous spend visibility, and in cost efficiency ^[5]. A 2025 Global CPO Survey points in the same direction: the CPOs investing in analytics and AI are the ones reporting durable value ^[6].