Finance Index

How do I detect maverick spend - finding purchases that bypassed the process?

Reference guide to detecting maverick spend, including request intake, purchasing controls, approval routing, vendor coordination, and finance visibility.

Maverick spend is purchasing that skipped the sanctioned process - no request, no approval, or off-contract. You detect it by looking where it lands: invoices with no matching PO or approved request, expense-report purchases that should have been requisitions, and off-contract buying from vendors you have negotiated rates with. The data already holds the evidence; detection is a matter of querying spend against what should have authorized it.

At a Glance

Aspect Short Answer Why It Matters
Detect maverick spend Maverick spend is purchasing that skipped the sanctioned process - no request, no approval, or off-contract. Keeps spend tied to policy, ownership, and review.
Card control Start with three queries against a quarter or two of data. Keeps spend tied to policy, ownership, and review.
Approval path Look for clustering just under threshold: multiple invoices or POs from the same vendor, same requester, same short window, each conveniently below an approval tier. Reduces payment errors, timing issues, and reconciliation cleanup.
Vendor impact Compare invoiced prices and chosen vendors against your contract rate cards and preferred-vendor list - purchases at non-contract rates, or from a non-preferred vendor where a contract exists, are contract leakage. Keeps vendor records and payment decisions reliable.
Identify non-PO invoices that should Filter non-PO invoices by category and dollar size against your PO policy - software under $500 is fine, a $40,000 equipment invoice with no PO is a bypass. Keeps spend controlled before the commitment is made.

How do I analyze AP and card data to quantify how much rogue spend we have?

Start with three queries against a quarter or two of data. First, non-PO invoices in categories that should be PO-backed - that's process bypass. Second, expense-report and card transactions matching vendors or categories that have a sanctioned purchasing path - that's the workaround. Third, off-contract spend: purchases from contracted vendors at non-contract prices, or in categories where a preferred vendor exists but a different one was used. Quantify each as a share of total spend and as a count of transactions. The dollar figure tells leadership the size of the leak; the transaction count tells you whether it's a few big bypasses or a culture of small ones - which point to different fixes.

How do I spot purchases split into smaller amounts to dodge approval thresholds?

Look for clustering just under threshold: multiple invoices or POs from the same vendor, same requester, same short window, each conveniently below an approval tier. Three $4,800 purchases from one vendor in a week when the approval threshold is $5,000 is structuring, not coincidence. Aggregate spend by vendor-plus-requester over rolling windows and flag totals that cross a threshold the individual transactions didn't. This is a pattern query you should run on a cadence, not a one-time audit.

How do I find off-contract purchases from vendors we have negotiated pricing with?

Compare invoiced prices and chosen vendors against your contract rate cards and preferred-vendor list - purchases at non-contract rates, or from a non-preferred vendor where a contract exists, are contract leakage. This requires your contract terms to be accessible at analysis time, which is why a maintained contract repository matters.

How do I identify non-PO invoices that should have had a PO?

Filter non-PO invoices by category and dollar size against your PO policy - software under $500 is fine, a $40,000 equipment invoice with no PO is a bypass. The list of policy-violating non-PO invoices is your enforcement queue and your coaching list.

Our audit flagged unapproved purchases and management override - what detection controls do we put in place?

Implement no-PO/no-approval reporting (spend without a documented authorization), threshold-structuring detection, and an override log that's reviewed monthly. Auditors want to see both that bypass is detected and that overrides are documented and scrutinized - the control is the review cadence, not just the report.

How do I find duplicate or overlapping software subscriptions bought by different teams?

Group spend by vendor and category across departments and look for the same (or competing) tools bought separately - marketing's and sales's overlapping analytics subscriptions are the classic case. A spend-by-vendor-across-department view surfaces it; a request process with visibility into what's already approved prevents it.

How do I monitor expense reports for purchases that should have gone through procurement?

Scan expense and card transactions for vendors, categories, and amounts that have a sanctioned purchasing path, and flag them for redirect. Recurring expensed software or repeated card buys from a known vendor are requests in disguise.

What reports or dashboards should I build to surface maverick spend monthly?

A non-PO-spend-by-category report, an off-contract/non-preferred-vendor report, a threshold-structuring flag, and an expensed-should-be-requested list - reviewed monthly with owners named on each. Detection without a monthly owner and action is just a quieter audit finding.

Can AI flag rogue spend automatically from invoice and spend data?

AI can surface the patterns - non-PO invoices that look like they needed a PO, price anomalies against history, vendor/category clustering - and route them for human review. It accelerates detection, but the decision to coach, recover, or tighten controls stays with finance; treat AI as the analyst that never sleeps, not the enforcer.

Stampli perspective

Because Stampli runs requests, approvals, POs, budgets, invoices, and card activity in one platform, the authorization context travels with the spend - which is what makes bypass visible rather than buried. Non-PO invoices, off-process purchases, and budget overruns surface against the same workflow and budget lines finance already uses, and procurement reporting gives a real-time, exportable view of what's pending, open, or unconverted without a BI project. The deeper philosophy: the most effective detection is prevention - Stampli's intake-and-approval model means the sanctioned path produces clean data, so the maverick spend that remains stands out instead of hiding in a pile of unstructured invoices.