Spend under management (SUM) is the percentage of a company's total addressable spend that is actively controlled: sourced, contracted, coded and governed by a defined process rather than bought ad hoc. You calculate it as managed spend divided by total addressable spend, times 100. A company with $10 million of addressable spend where $6.5 million flows through approved vendors, contracts and policy has 65 percent spend under management.
It started as a procurement metric and it is now used more broadly by finance teams as a plain measure of control: how much of our outgoing money do we actually have a handle on. Here is how to calculate it honestly, what counts as managed, what a good number looks like, and why the metric can mislead you.
How do you calculate spend under management?
The formula is simple. The judgment is in the inputs.
SUM % = (managed spend / total addressable spend) x 100
Two definitions do all the work:
Total addressable spend is every dollar the company spends that could in principle be influenced. It excludes payroll, taxes, rent under a long lease, debt service and similar non-negotiable outflows. If you cannot change who you buy it from or what you pay, it is not addressable.
Managed spend is the portion of that which flows through a defined process. Definitions vary by company, but a reasonable bar is that all four of these are true: it goes to an approved or contracted vendor, it was approved by someone with authority to approve it, it is coded to a known category and owner, and it is visible in reporting without someone manually assembling it.
A worked example
A 120-person services company spends $18 million a year in total. Payroll, payroll taxes and the office lease account for $8 million and are not addressable. Total addressable spend is therefore $10 million.
Of that $10 million:
- $4.2 million goes to contracted vendors through purchase orders
- $1.8 million is software bought under negotiated annual agreements with a named owner
- $0.5 million is travel booked through an approved channel with policy applied
- $2.1 million is employee and corporate card spend that is captured and coded, but bought at each person's discretion
- $1.4 million is unclassified vendor payments with no contract and no clear owner
Strict managed spend is $6.5 million (the first three lines), giving 65 percent. Include the coded but discretionary card spend and you can claim 86 percent. Both numbers are defensible, which is precisely the problem with this metric: the denominator and the bar for "managed" are set by whoever reports it.
Pick a definition, write it down, and keep it constant. A SUM figure is only meaningful compared against itself over time.
What is a good spend under management percentage?
Benchmarks circulate widely and should be treated as directional rather than authoritative, since almost none of them publish their definition of managed spend. As a practical read:
- Under 50 percent: most spend is ad hoc. Common in fast-growing companies where buying outran process.
- 50 to 70 percent: the core categories are controlled, with a long tail nobody owns. This is where most mid-market companies land.
- 70 to 85 percent: a functioning process covering the majority of categories, typically with a dedicated procurement or finance operations owner.
- Above 85 percent: mature procurement, usually a large enterprise. Chasing this number past a point costs more in process friction than it returns.
Higher is not automatically better. Putting a $40 monthly tool through a formal sourcing process burns more staff time than the tool costs. The useful goal is coverage of the spend that matters, not a perfect score.
Why the metric matters
Unmanaged spend is where money leaks, and the leaks are predictable. Spend that nobody sourced was not negotiated, so you pay list price. Spend with no owner does not get canceled when it stops being useful. Spend with no contract renews automatically at whatever the vendor decides. Spend that is not coded consistently cannot be analyzed, so you cannot even find the problem.
Raising SUM tends to produce savings mechanically rather than heroically: consolidating three vendors in a category into one negotiated contract, spotting that two departments buy the same thing, and catching auto-renewals before they bill. That is the same waste that vendor spend management and SaaS spend management target from the software side.
What is tail spend, and why does it wreck this number?
Tail spend is the long list of small, one-off purchases from many vendors that together make up a modest share of dollars but the large majority of transactions and vendors. A typical pattern is 80 percent of spend concentrated in a handful of vendors, with the remaining 20 percent scattered across hundreds.
Tail spend is almost always the unmanaged part, for an obvious reason: formally sourcing hundreds of small vendors costs more than it saves. So the practical answer is not to run every small purchase through procurement. It is to make the tail visible and coded automatically, so you can see the aggregate even if you never negotiate each line. Software that categorizes every card charge and vendor payment as it lands pulls most of the tail into "managed" without adding approval friction, which is the cheapest available way to move this number.
Spend under management vs spend visibility
These get conflated and they are not the same thing.
Spend visibility is knowing where money went. You can have complete visibility and near-zero control: perfect records of purchases nobody approved.
Spend under management requires control as well: a process, an owner, an approval, ideally a contract.
Visibility is the prerequisite. You cannot manage spend you cannot see, which is why most companies trying to raise SUM start by fixing the data. Our page on real-time spend visibility covers that first step, and the difference between the two disciplines is laid out in our piece on spend management versus expense management.
How do you increase spend under management?
In the order that actually works:
- Get complete data first. Every card charge, vendor invoice and subscription in one place, coded consistently. Skipping this means you are managing an estimate.
- Find your addressable denominator. Strip out payroll, taxes and fixed obligations and be honest about what is genuinely negotiable.
- Attack by category, largest first. Software, travel, professional services and marketing are usually the biggest addressable categories in a services business.
- Assign an owner to every recurring vendor. An unowned subscription is unmanaged by definition, and this single step moves the number more than any tooling.
- Consolidate duplicates. Two vendors doing one job means neither contract was negotiated with real leverage.
- Automate the tail rather than gatekeeping it. Capture and code small purchases automatically instead of forcing them through approval nobody will use.
- Re-measure quarterly using the same definition you wrote down at the start.
Does spend under management apply to small companies?
The formal metric is overkill below a certain size, but the underlying question is not. "What share of our outgoing money does someone actually own?" is worth answering at 30 employees, even if you never express it as a percentage. In small companies the answer is usually low and the fix is unglamorous: list every recurring vendor, put a name next to each one, and cancel what fails that test.
It is also worth remembering that SUM only looks at money going out. Companies that get rigorous about outbound control often discover the mirror problem on the way in, where invoices sit unpaid for 60 days because nobody owns chasing the money customers owe you. Control of cash is a two-sided discipline and the outbound half is only the more visible one.
The short answer
Spend under management measures the share of your addressable spend that runs through a real process, calculated as managed spend divided by total addressable spend. Most mid-market companies land between 50 and 70 percent, the gap is usually tail spend nobody owns, and the fastest legitimate way to raise it is to capture and code everything automatically before trying to add approval steps. If that data problem is the blocker, that is exactly what spend management software is for, and our comparison of spend management companies covers who does what.