Written by Marijn Overvest | Reviewed by Sjoerd Goedhart | Fact Checked by Ruud Emonds | Our editorial policy

How to Find Maverick Spend With Claude (Before Your CFO Does)

What Maverick Spend Actually Is (The Definition Gap Most Teams Have)

Maverick spend is the spend that bypasses the procurement function's preferred path. The preferred path is whatever the team has decided is the controlled route, an approved supplier, a negotiated contract, a defined catalogue, a routed approval workflow. Anything outside of that path is maverick. The definition sounds simple, but most procurement teams don't have a written one, which is why maverick spend keeps quietly growing.

Most procurement teams find that isolated experiments with Claude only become a durable team capability when tool practice is paired with structured training. The AI Fundamentals for Procurement Teams program is built for exactly that transition, from individual curiosity to a procurement function that works differently.

In practice, maverick spend takes five forms. Off-contract spend: buying from a supplier with whom there is no negotiated agreement, often at list price. Off-catalogue spend: buying items outside the approved catalogue when an equivalent item exists on it, usually at higher cost. Wrong-supplier spend: routing volume to a non-preferred supplier when a preferred one was available. Tail-spend leakage: hundreds of small transactions with one-off suppliers that, in aggregate, would justify consolidation. And urgent-buying spend: the requisitions raised after a stockout or service failure, with no time for sourcing, at premium prices the procurement team would never have agreed to under normal conditions.

The CFO sees maverick spend as a controllable cost leak. Procurement teams that can quantify it and show a credible plan to reduce it become a strategic conversation partner with finance. Teams that cannot risk being framed as a transactional function. The conversation matters; the data to support it usually doesn't exist until someone builds it.

Why Maverick Spend is so Hard to Detect in Practice

Three structural reasons that explain why most procurement teams have a vague sense of their maverick exposure but no defensible number.

Data fragmentation. The data needed to detect maverick spend lives in different systems. The list of preferred suppliers is in the procurement contracts repository or the eSourcing suite. The actual transactions are in the ERP or the P2P system. The approval workflow logs are in a third place. Joining them across a vendor master that may have three variants of the same supplier name is non-trivial, and is usually not part of the standard month-end reporting.

The vendor master problem. The same supplier appears in the AP system as "Acme Inc.," "Acme Incorporated," "Acme Inc" (no period), and "ACME-USA-2387." Each variant looks like a separate vendor in a naive analysis. A meaningful maverick analysis requires upstream normalisation of vendor names, which is a category-management capability most procurement teams don't have a tool for.

Definition drift. What counts as "off-contract" depends on the contract. A two-year MSA with an evergreen clause is still in force; a one-year agreement that lapsed in March is not. Without a clean view of which contracts are current, the "off-contract" definition keeps shifting, and the analysis loses credibility. Contract Management Course covers the contract hygiene foundations that make this kind of analysis possible.

The three problems are why maverick spend remained largely unanalysed for a generation of procurement teams. With Claude, the joining and normalisation work is fast enough to be done monthly instead of yearly, which is what changes the game.

The Five Maverick Spend Patterns Claude Can Find for You

When pointing Claude at a spend file, the goal is not "find maverick spend" as a single ask. The goal is to find each of the five distinct patterns, because each one has a different root cause and a different remediation.

Pattern 1, off-contract suppliers with significant volume

Cross-reference the supplier list in the spend file against the list of suppliers with current contracts. Anything spending more than a defined threshold, typically EUR/USD 25,000 or local currency equivalent annually, with no active contract is a candidate. These suppliers are either eligible for consolidation under an existing master agreement, candidates for a new agreement, or candidates for elimination.

Pattern 2, same supplier under multiple vendor IDs

Run a fuzzy-match across the supplier names in the spend file. Variants of the same supplier are typically the largest source of "hidden" volume aggregation opportunities. Total the combined spend; if it crosses the threshold, this supplier is in fact a strategic vendor that the data is hiding.

Pattern 3, off-catalogue purchases with an on-catalogue equivalent

Compare item descriptions in the spend file against the approved catalogue. Items that match by description but were purchased off-catalogue, at higher price points, indicate either a user-experience problem with the catalogue (people couldn't find what they needed) or a compliance problem (people skipped the catalogue deliberately).

Pattern 4, urgent-buying premiums

Look at lead times or requisition-to-PO times in the data. Requisitions cycled in under a defined threshold (24 hours is common) are likely urgent buys. Compare prices on urgent buys against the same supplier's standard-lead-time prices. The delta is the urgency premium, and it accumulates across the year.

Pattern 5, tail-spend clustering

Identify suppliers below a small-supplier threshold (under EUR/USD 10,000 annual spend, for example). Cluster them by category. Clusters of 20+ small suppliers in a single category usually represent a consolidation opportunity, the work to negotiate a single supplier or a small panel typically pays back inside a quarter.

The Claude Detection Workflow: From Raw Spend to Ranked Findings

Here is the practical workflow procurement teams use to run a monthly maverick check. Total time once the workflow is set up: about 45 minutes. The setup is a one-off afternoon.

Setup, the one-off afternoon. Create a Claude Project called "Maverick Spend Analysis." Upload three reference documents: the list of preferred suppliers with their categories and contracts, the approved-catalogue items list, and the team's policy thresholds (what counts as off-contract, what counts as urgent). Add instructions in the Project: "This Project analyses spend data for maverick patterns. Always run all five patterns. Output findings ranked by exposure value. Flag uncertainty for patterns where the data is ambiguous."

Monthly run, step 1. Export the latest month's spend transactions from the ERP or P2P system into CSV. Keep the columns the analysis needs: vendor name, vendor ID, GL or category code, item description, gross amount, PO date, requisition date, requester. Upload into the Project.

Monthly run, step 2. Invoke the analysis: "Run the five-pattern maverick check on the attached spend. Return findings as a ranked list with: pattern, supplier, total exposure, suggested action, evidence reference. Flag any pattern where confidence is low."

Monthly run, step 3. Review the output with the category lead responsible for the area where the maverick is concentrated. Decide which findings to action this month and which to log for the quarterly review. Update the action register; the action register is the procurement team's own follow-through, Claude finds the patterns but cannot enforce category strategy.

Monthly run, step 4. Track the recovered value over the following months. The point of the analysis is not the report; it is the closed-loop where the team confirms whether maverick spend reduced after each intervention. Without that loop, the analysis becomes another procurement deliverable that nobody acts on.

The Prompt Template for Monthly Maverick Analysis

The prompt below is the working template procurement teams have used. Adapt the thresholds and definitions to your organisation.

"You are running a monthly maverick spend analysis on the attached spend file. The team's preferred-supplier list, approved catalogue, and policy thresholds are in the Project context. Run the five patterns:

• Pattern 1: Off-contract suppliers with annual spend over [EUR 25,000]. List with total spend and the most likely existing or candidate contract for consolidation.
• Pattern 2: Same supplier under multiple vendor IDs. Use fuzzy matching, surface only matches with combined spend over [EUR 50,000]. List the variants and the consolidated total.
• Pattern 3: Off-catalogue purchases of items with a likely on-catalogue equivalent. Compare item descriptions; surface matches over [EUR 10,000] per item.
• Pattern 4: Urgent purchases (requisition-to-PO under [24] hours) with the same supplier's standard-lead-time prices available. Calculate the urgency premium per finding.
• Pattern 5: Tail-spend clusters: categories with [20+] suppliers each under [EUR 10,000] annual spend. List the clusters with the count and total category exposure.

For each finding, output: pattern number, supplier or cluster, exposure value, confidence level (high / medium / low), recommended action (consolidate / sourcing event / catalogue addition / urgency root-cause review). Rank by exposure value. Flag patterns where you needed to make assumptions and what assumptions you made."

The prompt is deliberately specific. Vague prompts produce vague output. Specific prompts produce work the category lead can act on the same week.

From Detection to Prevention: Closing the Loop

Detection without prevention is theatre. Three actions, each tied to one of the patterns, that turn the monthly analysis into a sustained reduction in maverick spend.

For Pattern 1 (off-contract suppliers), set a quarterly consolidation target. Each quarter, take the top three off-contract suppliers by exposure. For each, run a quick contract conversation: bring under an existing master, negotiate a new agreement, or eliminate. Track the recovered spend over the next two quarters; this becomes the procurement team's defensible savings number.

For Pattern 3 (off-catalogue purchases), do a quarterly catalogue refresh. The top off-catalogue items often indicate gaps in the approved catalogue, items people needed but couldn't find. Adding them to the catalogue at the preferred-supplier price typically closes 60-80% of off-catalogue volume in that category within two months. Category Management in Procurement Course covers this catalogue-curation discipline in depth.

For Pattern 4 (urgent-buying premiums), do a root-cause workshop. Urgent buys are rarely about procurement; they're about demand planning, stockouts, or unplanned project work. Sharing the urgency-premium data with the requesting business units, framed as "this is what we paid for being late," changes the conversation. The fix usually lives outside procurement, but procurement has to start the conversation.

Quantifying the Savings: The Conversation with Finance

The maverick spend program lives or dies by the credibility of its savings number with finance. Three principles that procurement teams use to make the conversation work.

Use evidence, not estimates. A maverick spend program that claims 15% savings on "identified maverick volume" without showing the contract delta or the consolidation evidence is not credible. A program that claims 12% savings on EUR 4M of consolidated volume, with the before-and-after pricing visible per supplier, is credible. Finance trusts numbers they can trace; build the evidence trail.

Separate one-off from recurring. Renegotiating a one-off off-contract supplier saves once. Adding the supplier to a multi-year agreement saves recurring. Finance values recurring savings 3-5x more than one-offs in the savings narrative. Frame the analysis with this distinction; don't blend them.

Show the program's marginal cost. The maverick spend program costs the procurement team's time. If it saves EUR 800K but costs three FTE-months of analyst and category-lead time, the ROI is still significant, and finance respects procurement teams that disclose their own cost rather than only claiming savings. Procurement teams that demonstrate financial impact credibly are the minority. The 2026 AI Readiness in Procurement survey found that 32% of procurement teams describe their financial impact as difficult to quantify for finance. A repeatable maverick spend program is one of the cleanest ways to move out of that group.

Common Pitfalls that Make Maverick Spend Programs Stall

Running the analysis once and stopping

A one-off maverick analysis is interesting; a monthly one changes behaviour. The single most common failure mode is that the procurement team runs the analysis in Q1, presents to finance, and then doesn't repeat it in Q2 because something more urgent came up. The compounding value disappears.

Treating every finding as a savings opportunity

Not every off-contract supplier is worth consolidating. A small supplier with a niche capability, a relationship with a specific business unit, or a tactical reason for the off-contract status may be best left alone. The skill in maverick spend is sorting findings into "act," "investigate," and "accept"; treating all findings as actionable burns goodwill with the business units.

Letting the analysis become procurement-only

The remediations live with the category leads, the requesting business units, and sometimes finance. A maverick spend program that the procurement team runs in isolation, without engaging the business units the spend is happening in, produces analysis but not change. The monthly findings should be reviewed jointly with the relevant business unit owners.

Underestimating the change-management dimension

Maverick spend often reflects a deeper issue: catalogues that don't match real demand, approval workflows that are too slow for urgent buys, or business units that don't trust procurement's preferred suppliers. The analysis surfaces the symptoms; the actual fix often requires changing how procurement operates with the rest of the business. Be ready for that scope of work, not just the analytical work.

Frequently asked questions

How much maverick spend is typical?

5-15% of total spend in most organisations. Varies by procurement maturity.

Can Claude quantify the savings opportunity?

Approximately, based on rate differentials between off-contract and preferred-supplier pricing.

What's the typical recovery rate from maverick spend programmes?

1-3% of total spend over 12-18 months of systematic attention.