Spend analysis
Collecting, cleaning and classifying all of a company's spending data to see what it buys, from whom and at what price.
In short
- It combines data from accounts payable, cards and expense reports.
- The hard part is cleaning: merging duplicate supplier names and classifying spend consistently.
- It is the starting point for sourcing, savings and spend control.
Spend analysis answers basic questions that are surprisingly hard to answer in most companies: how much do we spend, on what, with which suppliers, by which departments, and are we getting the prices we negotiated?
The steps
- Collect: gather spending data from every source: accounts payable, purchase orders, corporate cards, expense reports and sometimes several ERPs.
- Clean: merge supplier name variations ("IBM", "I.B.M. Canada", "IBM Corp") into one supplier, and fix obvious errors.
- Classify: assign each transaction to a spend category, usually using a standard taxonomy, often with the help of line-item detail and merchant category codes.
- Analyze: look at spend by category, supplier, department and trend over time.
What it reveals
- The same items bought from many suppliers at different prices
- Spend with suppliers who have no contract
- Duplicate subscriptions and overlapping tools
- Categories no one owns
- Opportunities to consolidate volume for better terms
A one-off analysis is useful, but its value fades as new transactions arrive. Keeping spend classified continuously, as it is captured, turns spend analysis from a project into a report anyone can open.
Worked example
What one analysis found
A company with $32 million of annual spend runs its first full analysis across AP, cards and expenses. It finds 2,400 active suppliers, 180 of which account for 80% of spend, and 14 different suppliers of cleaning services across its sites. Consolidating cleaning to two regional suppliers saves 12% on a $1.1 million category.