Carrying out a spend analysis: from data dump to savings agenda

Ask three departments how much the organisation spends with a particular supplier and you get three answers. Procurement looks at contracts, finance at general ledger accounts and the business at budgets. A spend analysis brings those realities together into one factual picture: what do we really spend, with whom, on what and on what terms. Only with that picture does saving become a targeted activity instead of across-the-board cuts.

Step 1: collect the right data

The basis is the payment history from the accounts payable system over at least 12, preferably 24 months: supplier, invoice amount, date, general ledger account, cost centre and, where available, purchase order and item data. Add the vendor master file and the main contracts. Perfection is not needed; 80 percent completeness is enough to get started.

Step 2: clean and classify

Merge duplicate suppliers (the same party under several vendor numbers distorts any picture) and classify the spend using a category tree: main categories such as logistics, IT, facilities and professional services, with two levels below. Classify automatically by supplier and ledger account first, and manually only the largest remaining items.

Step 3: analyse for the classic patterns

• Fragmentation: many suppliers in the same category point to bundling potential and better volume tiers.
• Maverick buying: spend outside contracts and purchase orders, recognisable by invoices without an order.
• Price drift: rising average prices at constant volumes, or surcharges that have stuck.
• Tail spend: the long tail of small suppliers that generates high processing costs and little value.
• Payment terms: differences between contractual and actual terms, and missed early-payment discounts.

Step 4: turn it into a savings agenda

An analysis without an agenda is a report for the drawer. Prioritise each category on two axes: savings potential and feasibility (contract term, dependency, internal resistance). The result is a concrete list: which categories do we renegotiate this quarter, which do we bundle, what do we terminate and which process agreements do we make. Give every action an owner and a deadline and report quarterly on savings actually realised.

Common mistakes

• Polishing the data for too long instead of starting with what is there.
• Reporting savings on the negotiated rate instead of actual realised spend.
• Using the accounts payable data only for the spend analysis: the same dataset can immediately be reviewed for duplicate payments and VAT errors as well, and those proceeds often fund the whole project.