The short answer
A spend analysis is a structured review of everything an organization buys: who it pays, what for, through which teams, and at what price. To do one, gather twelve months of invoice and payables data, clean it so each supplier appears once under one name, classify every line into spend categories, and then test the result for the five patterns that cost money: spend concentration, price variance, vendor fragmentation, off-contract buying, and duplication and waste.
The output is not a dashboard but a list of sized opportunities, each traceable to the invoices behind it, with an owner and a date. That is what turns an analysis into procurement savings a board will approve.
What a spend analysis is, and why a year of data
The ledger records what was paid. It does not group the same supplier under one name, say what was actually bought, or compare one price with another. A spend analysis builds that view by answering three questions for every line: who was paid (the supplier, rolled up to its parent company), what was bought (the category), and who bought it (the business unit, site or cost center). Once spend can be sliced on all three, most savings questions become simple queries.
Twelve months is the usual window. It covers seasonal buying and annual renewals, and gives enough volume per item to compare prices reliably. Much longer windows mix in prices and contracts that no longer apply.
How to do a spend analysis, step by step
1. Gather the data
Start with the accounts payable ledger or invoice register for the last twelve months, then add what explains it:
- Invoice lines: supplier, number, date, amount, currency, description, and quantity and unit price where they exist.
- Purchase orders, to link each invoice to who requested it and whether it was approved in advance.
- Contracts: which suppliers are under contract, for what, at which prices and until when.
- Card and expense data, where much of the small, untracked buying happens.
Reconcile the total to the general ledger before going further. If the data does not add up to what finance reports, every later number will be questioned.
2. Clean and normalize suppliers
The same company usually appears under several names: abbreviations, misspellings, subsidiaries, old names kept after a merger. Until those are grouped, spend with the largest suppliers is understated and fragmentation overstated. Normalizing means matching name variants, tax identifiers and bank details to one supplier record, linking subsidiaries to their parent, converting amounts to one currency at a consistent rate, and aligning units of measure.
3. Classify spend into categories
Each line is then mapped to a category taxonomy, typically two or three levels deep: facilities, then cleaning services, then janitorial contracts. Ledger codes are a weak starting point, because they record how a cost was booked, not what was bought; line descriptions and purchase order text do most of the work. Published standards such as UNSPSC exist, but many firms use a simpler taxonomy that matches how their buyers are organized.
Classify by value, not by row. A small number of suppliers usually holds most of the spend, so getting those right first covers most of the value quickly.
4. Analyze
With clean, classified data, run the five tests below across every category.
The five places savings hide
| Where savings hide | How to find it | Typical fix |
|---|---|---|
| Spend concentration | Rank suppliers by spend, overall and per category. Flag categories held by one or two suppliers, and any single-source dependence. | Use the volume as leverage in a renegotiation or tender; add a second source where dependence is a risk. |
| Price variance for the same item | Compare unit prices for the same item across sites, teams, suppliers and quarters, after aligning units. | Move volume to the best price already achieved, put the item on a price agreement, and recover overcharges. |
| Vendor fragmentation | Count suppliers per category. Many suppliers each doing a small share of the same job is the pattern. | Consolidate onto a few preferred suppliers for volume pricing and less overhead. |
| Off-contract buying | Match contracted categories against the contract list: buying from suppliers without a contract, or from contracted ones at other than agreed prices. | Route buying through the contract with catalogs, purchase order rules and approvals. |
| Duplication and waste | Search for invoices paid twice, purchases paid through two channels, and subscriptions or services still billed after use ended. | Recover duplicate payments, cancel what is unused, and add duplicate checks before payment. |
Two cautions. The five overlap: a fragmented category usually shows price variance too, so the same spend can surface twice. And not every variance is waste: a higher price may buy faster delivery or a second source the business needs. Each finding needs a buyer's review before it becomes a target.
Sizing savings conservatively
A savings case is only as strong as its weakest number, so size each opportunity the way a skeptical finance director would.
- Start from addressable spend. Exclude what cannot change in the period: taxes, regulated fees and contracts locked in beyond the horizon.
- Use prices already achieved. The target price should be the best price the organization already pays somewhere, not an outside benchmark nobody can verify.
- Apply a realistic capture rate. Size only the volume that can actually shift within the year.
- Count each saving once, under the lever most likely to deliver it.
- Separate one-off from recurring. A recovered duplicate payment is cash once; a better unit price saves every year.
- Time it to contracts, and net off the cost of capturing it: switching costs, tooling, staff time.
A hypothetical example shows the difference. A company buys 20,000 units of one item a year at an average of $11.50, while one site already pays $10.00 under contract. The theoretical saving is 20,000 × $1.50 = $30,000. If only the sites buying 40% of the volume can switch this year, the conservative figure is 8,000 × $1.50 = $12,000. That is the number that goes in the case, with $30,000 shown as the potential once every site has moved.
Building a board-ready savings case
A board does not approve an analysis; it approves decisions. A good savings case is short at the top and deep underneath.
- One total, as a range: the conservative figure and the full potential, split into recurring savings and one-off recoveries.
- The breakdown by lever and by category, so the board sees where the money is and which teams it involves.
- The top opportunities, each with its size, evidence, owner, timeline and the decision needed.
- A trail to the source. Every finding should open down to the invoices, suppliers and prices behind it. When a director asks where a number comes from, the answer should be the firm's own data, not an industry average.
- The risks: supplier dependence, quality trade-offs, and anything that could delay delivery.
If the spend analyzed agrees with the general ledger and every finding traces back to the lines that produced it, the debate moves from whether the numbers are right to which savings to pursue first.
Keeping it current
A one-off analysis decays quickly: new suppliers appear, contracts expire and buyers drift off contract again. Treat spend analysis as a recurring discipline.
- Refresh on a schedule, monthly or quarterly, from the same source systems, keeping the supplier matches and category mappings already made.
- Track realized against identified, so the next case starts with credibility.
- Watch compliance. Off-contract buying and duplicate payments are easiest to fix in the month they happen.
- Connect to the source. Pulling data straight from the ERP and procurement systems removes the manual exports that make each refresh slow.
How OUTLAY does it
OUTLAY: a year of invoices, turned into a savings case
OUTLAY is enterprise spend analysis that turns a year of invoices into a board-ready savings case. It classifies the spend, quantifies each opportunity and shows what can be recovered, with every finding traced to your own data and sized in your currency.
It looks in the five places this guide describes. Spend concentration: who you actually pay, how exposed you are to them, and where a handful of suppliers hold leverage that should be yours. Price variance: the same item bought at many prices by different teams, sites and quarters. Vendor fragmentation: dozens of suppliers doing one job, with the lost volume discounts, duplicated overhead and weak negotiating ground that come with it. Off-contract buying: purchases made outside the deals you already negotiated. Duplication and waste: paying twice, paying for the unused, paying the same invoice through two doors.
OUTLAY integrates natively with SAP S/4HANA, SAP Ariba, Oracle NetSuite and Coupa.
Questions
What is a spend analysis?
A spend analysis is a structured review of everything an organization buys, showing who it pays, for what, through which teams and at what price. Its purpose is to find specific, sized savings.
How do you do a spend analysis?
Gather twelve months of invoice data with contracts and purchase orders, clean it so each supplier appears once, and classify every line into categories. Then test each category for concentration, price variance, fragmentation, off-contract buying, and duplication and waste.
What is off-contract spend?
Off-contract spend is buying outside the agreements an organization has already negotiated, either from suppliers without a contract or from contracted suppliers at other than agreed prices. It gives up discounts the organization has already earned.
What is vendor consolidation?
Vendor consolidation means moving the spend in a category from many suppliers to a few preferred ones. It usually brings better volume pricing and less overhead, weighed against the risk of depending on too few.
How do you estimate procurement savings without overstating them?
Start from spend that can actually change, use prices the organization already pays elsewhere as targets, and count only the volume that will realistically move within the year. Count each saving once and report one-off recoveries separately from recurring savings.
