What Spend Under Management Means
Spend under management (SUM) is the proportion of an organisation’s total addressable spend that procurement actively controls — through competitive sourcing, contracts, preferred suppliers and compliant buying channels — expressed as a percentage of the spend it could influence. It is the single clearest gauge of how much of a company’s purchasing is governed rather than left to discretion.
The concept matters because uncontrolled spend is, almost by definition, expensive spend. A purchase made off-contract, from an unvetted supplier, at a price nobody negotiated, leaks money and creates compliance and audit risk. SUM puts a number on how much of that leakage a procurement function has closed. It is one of the metrics a CPO is most often asked to defend, which is why it features prominently in our complete list of procurement KPIs and in the strategic priorities set out in the procurement AI guide for the CPO.
Key takeaways
- SUM = actively managed spend ÷ total addressable spend × 100.
- Typical well-run functions sit around 70–85%; below 60% signals significant uncontrolled spend.
- The formula is trivial; the judgement — what counts as “managed” and “addressable” — is where the work is.
- AI raises SUM mainly by reclaiming tail spend, fixing classification, and steering buyers to compliant channels.
The Formula — and Its Two Hard Parts
The calculation is simple arithmetic:
SUM % = (Actively managed spend ÷ Total addressable spend) × 100
The difficulty lives entirely in the two inputs. Total addressable spend is all spend procurement could realistically influence. It deliberately excludes categories the function cannot touch — taxes, statutory payroll, intercompany transfers, certain regulated or legally mandated costs. Get this denominator wrong and the ratio is meaningless: inflate it with non-addressable spend and SUM looks artificially low; shrink it and SUM looks flattering but hollow.
Actively managed spend is the numerator and the more contested figure. It is purchasing routed through a procurement-controlled mechanism. The honest question every team must answer is what genuinely qualifies, because the temptation to count loosely is strong and corrodes the metric’s value.
What Actually Counts as “Managed”
Spend is reasonably treated as managed when it meets at least one substantive control test:
- Competitively sourced — awarded through an RFx or negotiation within a defined period (often the last 24–36 months).
- Under active contract — governed by current negotiated terms, not an expired or evergreen agreement nobody owns.
- Through a preferred supplier or compliant catalogue — bought via a channel procurement has set up and steers buyers toward.
- Category-managed — owned by a category manager with a live strategy, not merely “visible” in a report.
The crucial distinction is between visibility and management. Seeing spend in a dashboard is not managing it. A category that is fully classified and reported but where buyers still purchase ad hoc from whoever they like is visible, not managed. Conflating the two is the most common way SUM figures get quietly overstated — and it is exactly the gap our guide on finding savings with AI spend analytics is designed to close.
“Visibility is knowing where the money goes. Management is changing where it goes. Spend under management measures the second, not the first — and the difference is usually several percentage points of savings.”
Realistic Benchmarks
There is no universal target, because the right number depends on sector, spend mix and maturity. Direct-spend-heavy manufacturers structurally run higher SUM than indirect-heavy service or technology businesses, because direct materials are inherently contracted and engineered. With that caveat, the following bands are a defensible orientation rather than audited fact.
| SUM range | What it usually indicates | Typical profile |
|---|---|---|
| Below 50% | Largely uncontrolled spend; early-stage function | Decentralised, little category strategy |
| 50–65% | Core categories controlled, long tail loose | Developing function, indirect gaps |
| 70–85% | Well-run function with most spend governed | Mature, category-managed |
| 85–90%+ | Highly governed, tail spend tackled | Advanced, often direct-heavy |
Two warnings about benchmarking. First, a high SUM achieved by defining the denominator narrowly is not a real achievement. Second, chasing the last few points has diminishing returns — pushing from 88% to 92% often costs more in effort than the incremental control is worth, where the same effort spent earlier in the curve would have paid back handsomely. Where SUM sits across the market is one of the structural signals we track in the State of Procurement AI 2026 report.
Why SUM Drives Savings, Compliance and Risk
SUM is watched because three outcomes move with it. Savings: negotiated price reductions, payment-term improvements and consolidation can only be realised on spend procurement actually touches, so SUM is the base that savings programmes multiply against. Compliance: managed spend flows through controlled channels with audit trails; unmanaged spend is where off-contract pricing, unapproved suppliers and policy breaches hide. Risk: a supplier nobody vetted, paid through a channel nobody governs, is precisely the exposure that surfaces in an incident.
This is why raising SUM is rarely an end in itself — it is the enabler. A point of additional SUM is a point of additional spend on which every other procurement lever can now operate. The relationship between SUM and realised value is part of the business-case logic in our CPO strategic guide.
The Tail Is Where SUM Goes to Die
In most organisations, the long tail of low-value, high-volume, fragmented purchases is the single biggest drag on SUM. It is uneconomic to source manually — the effort to run an RFP on a $4,000 category exceeds any plausible saving — so it is left ungoverned, and it accumulates. Tail spend can represent a small fraction of value but a large fraction of transactions and suppliers, dragging the SUM percentage down even when the strategic categories are well managed.
Closing this gap is its own discipline, covered in depth in our complete guide to tail spend. The headline is that the tail was historically the part of SUM you simply wrote off — and it is the part that automation has most changed.
How AI Raises Spend Under Management
AI improves SUM through three distinct mechanisms, each attacking a different cause of leakage:
1. Classification that reveals hidden managed spend
A surprising share of “unmanaged” spend is simply miscoded. AI spend classification — from tools tracked in our spend analytics AI category such as Sievo and SpendHQ — reclassifies transactions to a consistent taxonomy, often surfacing spend that was contracted all along but invisible to the old coding. Part of the SUM gain is reclaiming credit for control you already had.
2. Guided buying that converts maverick spend
Intake and guided-buying tools steer requesters to compliant channels at the moment of purchase, turning would-be maverick buying into governed buying. Suite copilots such as those in Coupa nudge users toward preferred suppliers and catalogues before an off-contract PO is ever created — raising SUM at the source rather than reporting on the leak afterwards.
3. Automated tail-spend sourcing
Autonomous sourcing makes the tail economical to govern for the first time, sourcing thousands of small events that no buyer could touch manually. This is the mechanism that has moved SUM ceilings upward in mature functions, because it brings the historically un-manageable tail inside the perimeter.
None of these is a silver bullet, and all depend on clean underlying data. But together they explain why SUM, long stuck for many organisations, has become movable again. For the buyer’s-eye view of choosing tools to do this, the best spend analytics tools for CFOs is a useful companion.
A Practical Sequence to Improve SUM
Improving SUM is a programme, not a one-off measurement. A defensible order of operations:
- Fix the denominator. Agree what is genuinely addressable so the baseline is honest.
- Classify everything. Get a clean, consistent spend taxonomy before claiming any number.
- Reclaim miscoded managed spend. Often the fastest, cheapest gain.
- Steer new spend to compliant channels. Guided buying stops the leak prospectively.
- Automate the tail. Bring the historically ungoverned long tail inside the perimeter.
- Re-measure honestly. Hold the definitions steady so the trend is real.
Done in that order, SUM improvement compounds: each step makes the next cheaper. Done out of order — chasing the tail before classification is clean, or claiming a number before the denominator is honest — it produces a figure that looks good and means little.
Frequently Asked Questions
What is spend under management?
It is the share of total addressable spend that procurement actively controls through sourcing, contracts, preferred suppliers and compliant channels, expressed as a percentage. The higher it is, the more of the organisation’s purchasing is governed rather than ad hoc.
How do you calculate it?
Managed spend divided by total addressable spend, times 100. The arithmetic is trivial; the judgement about what counts as managed and what is addressable is where credibility is won or lost.
What is a good percentage?
Roughly 70–85% characterises a well-run function; below 60% signals meaningful uncontrolled spend; above 85–90% reflects a mature, highly governed operation. Direct-heavy sectors run higher than indirect-heavy ones.
How does AI increase it?
By reclassifying hidden managed spend, steering buyers to compliant channels through guided buying, and making tail spend economical to source — the three levers that move a previously stuck number.
Is high SUM always good?
Only if the definitions are honest. A high figure produced by a narrow denominator or by counting visible-but-not-managed spend is a vanity metric. Real SUM improvement shows up as savings, compliance and reduced risk, not just a better slide.