Written and maintained by CASRAI Editorial Board
Last updated
Short answer
Spend analysis is the practice of working out how much an organization spends on which goods and services, who it buys them from, and who is doing the buying. The US Government Accountability Office (GAO) describes it as a tool that gives an organization knowledge about “how much is being spent for what goods and services.” For a laboratory, clinic or hospital supply operation, it turns scattered purchase records into one comparable picture of consumables, equipment and services.
Definition and scope
GAO’s 2004 report on best practices, Using Spend Analysis to Help Agencies Take a More Strategic Approach to Procurement (GAO-04-870), frames spend analysis around three questions: what was bought, from whom it was bought, and who was purchasing it. The answers are the starting point for every later purchasing decision, because they show where money actually goes before anyone tries to change it.
Spend analysis is a view of past and current purchasing. It is not a purchasing method in itself. How an organization then organizes buying around that view is covered by category management, and the routes used to buy are covered in GSA schedule versus open-market purchasing. The analysis comes first; the buying strategy follows from what it shows.
How spend analysis is done in practice
GAO studied private-sector practice and summarized it as five processes: automating, extracting, supplementing, organizing and analyzing data. It describes them as ongoing work rather than a one-off exercise. The table below sets GAO’s wording beside what each step means when the dataset is laboratory or clinical purchasing.
| GAO process | What GAO describes | In a lab or clinic purchasing dataset |
|---|---|---|
| Automating | Spend data is compiled automatically rather than assembled by hand each time. | Pull purchase records on a schedule instead of building a spreadsheet for each review. |
| Extracting | Essential data is culled from accounts payable and other internal systems. | Take supplier, item description, quantity, amount and buying unit from invoices, purchase orders and card statements, including procurement card transactions. |
| Supplementing | Additional data is sought from other internal and external sources. | Add information the payment record lacks, such as a standard product name or a supplier’s parent company. |
| Organizing | The data is reviewed for accuracy and completeness. | Merge duplicate supplier names and identical items listed under different descriptions. |
| Analyzing | The data is analyzed continually with standard reporting and analytical tools. | Report spend by item group, supplier and buying unit, repeated on a regular cycle. |
Why the organizing step matters most
GAO found that organizations’ lists of items and suppliers often contained entries “that in reality are identical to each other,” that some organizations were unsure who their buyers were, and that databases lacked detail on the goods and services purchased. GAO also raised concerns about the reliability of the federal procurement data it examined. When the same box of gloves appears under three descriptions and two supplier spellings, the total for that item is split across rows and no single line looks significant.
Scope decisions: what the dataset leaves out
Every spend analysis starts with choices about what is in the data. A 2003 GAO report on a Department of Defense (DOD) pilot of spend analysis for services acquisition (GAO-03-661) shows how explicit those choices can be. The pilot drew on data from two databases, excluded research and development services, and did not include contract actions of $25,000 or less. GAO also noted there was no single financial data system for procurements. GAO also noted that the pilot’s data automation was a one-time requirement rather than a repeatable process, and that DOD said it intended to automate collection and analysis to make it repeatable.
Those limits give a checklist worth running on any purchasing dataset:
- Which systems feed it? Spend recorded in a system that is not connected does not appear.
- What is excluded by category? The DOD pilot left out a whole service type.
- Is there a size threshold? A cutoff such as $25,000 silently removes every smaller transaction, and for consumables bought in many small orders that can remove most of the lines.
- Is it repeatable? GAO noted the DOD pilot’s data automation was a one-time requirement, not the repeatable process private companies use; a one-time extract goes stale, while an ongoing process stays comparable.
Where it matters for labs, clinics and supply decisions
GAO’s report includes a medical example. The Department of Veterans Affairs (VA) was, at the time of the report, beginning to develop a spend analysis tool for medical and surgical supplies and high-technology medical equipment. It was working with a contractor to create uniform medical-product names, and planned to extract supply and equipment procurement data automatically into a central data warehouse. Commodity teams were pursuing national contracts on 45 categories of high-technology, high-cost items. The tool was not planned to cover facilities’ clinical care and support services procurement data.
The lesson for any supply operation is the order of work: product naming and data consolidation come before contract strategy, because a contract team cannot target a category it cannot count. The same logic applies to items whose selection is driven by clinicians, discussed in the guide to physician preference items, and to purchasing through a group purchasing organization, where the organization still needs its own record of what it buys.
Common misunderstandings
“Spend analysis is a software product”
GAO describes it as a set of processes applied to data. Software can carry out the steps, but the definition is the practice of knowing what is spent, with whom and by whom.
“The savings figures in the GAO report are medical supply savings”
They are not all the same kind. In GAO-04-870, the $394 million fiscal year 2003 figure for the VA relates to its strategic approach to pharmaceutical procurement, not to supplies or equipment. The $82 million a year figure is a potential saving that an inspector general estimated could come from more national contracts, not a realized result. These figures date from 2004 and should not be read as current outcomes.
“Spend analysis guarantees savings”
GAO reports that, in private-sector experience, spend analysis led to cost savings for some companies of 10 to 20 percent of their total procurement costs. That is a report of what some companies experienced, not a promise or a benchmark for any particular organization.
“It is a one-time audit”
Both GAO reports treat it as continuing work. GAO contrasted the DOD pilot’s one-time data requirement with the repeatable process used in private companies, and it lists continual analysis as one of the five processes.
FAQ
What is the difference between spend analysis and category management?
Spend analysis is the measurement step: it shows what is bought, from whom and by whom. Category management is how an organization then manages groups of similar purchases, for example by setting spend-tier expectations and coordinating buying. The first supplies the data the second relies on. See the CASRAI guide to category management for the management side.
Sources
- US Government Accountability Office, GAO-04-870, Best Practices: Using Spend Analysis to Help Agencies Take a More Strategic Approach to Procurement (September 2004). govinfo.gov/content/pkg/GAOREPORTS-GAO-04-870. Checked 2026-10-29.
- US Government Accountability Office, GAO-03-661, report on the Department of Defense services acquisition spend analysis pilot (June 2003). govinfo.gov/content/pkg/GAOREPORTS-GAO-03-661. Checked 2026-10-29.
