Recurring SaaS, duplicates, and shadow IT: the audit Procurement never runs often enough
The same vendor can be paid under 3 or 4 different supplier names, by teams that never talk to each other. Claude spots these duplicates and scattered AI spend in one prompt.
The problem
In a 250-person European tech company, SaaS subscriptions almost never go through a single purchasing channel. One employee opens a card to test a tool, a manager approves another for their team, and six months later nobody knows how much the company actually pays for a given tool category — or whether two subscriptions are, in fact, funding the same vendor under two different names.
This isn't mismanagement — it's simply that no consolidated view exists by default. A Procurement lead auditing the SaaS fleet before a renewal negotiation has to cross-reference exports, eyeball duplicates, and hope nothing was missed on AI spend — the category that has grown fastest over the last two years.
Claude, connected to the Spendesk MCP, does this reconciliation work directly on the company's real data.
What the Spendesk MCP changes with Claude
A recurring spend ranking, cross-referenced with cardholders, is enough to surface the duplicates and spend clusters that matter.
I analyzed recurring (subscription-type) spend over the last 12 months.
Top recurring suppliers
| Supplier | Annual amount | Share of total |
|---|---|---|
| Zendesk | €142,300 | 15% |
| Salesforce | €118,900 | 12% |
| Slack | €94,200 | 9% |
Want me to break down the exact amounts on each record before a possible merge?
Going further: 2 additional use cases
Mapping AI tool spend
Useful for a budget review or an AI tool usage policy.
Want me to group these lines into a single "AI Spend" view for your next budget review?
Spotting subscriptions opened without central coordination
Useful for a first-pass shadow IT check, before going further on compliance.
These three subscriptions aren't necessarily problematic, but none appears in a visible purchasing process from the MCP — worth checking whether a purchasing policy already covers them.
Want me to check if there are already-contracted alternatives for these use cases?
Key takeaways
- The same vendor can appear under several supplier names — Claude detects this by cross-referencing amounts and labels, but merging remains a manual action.
- AI spend is rarely natively consolidated — it has to be reconstructed from supplier names and active subscriptions.
- Subscriptions opened individually, with no pre-referenced supplier, are the first shadow IT signal worth checking.
- This audit is a starting point, not automatic compliance: every suspected duplicate must be confirmed before any action.
- Run it before every renewal negotiation or quarterly budget review.
Related skills
Identify suppliers representing an abnormal share of total spend and assess dependency risk.
Apply Pareto analysis to supplier spend, identify the long tail of micro-suppliers, and surface consolidation opportunities.
Cartographier les dépenses SaaS récurrentes, repérer les fournisseurs potentiellement dupliqués et isoler le spend IA pour détecter le shadow IT