Your Procurement 'Savings' Are Identified, Not Enforced

Start by giving the category its full due, because it earns it. Strategic sourcing and negotiation genuinely find value — a competitive event, a consolidated volume, a better-priced tier really do move the unit cost before anyone signs. Spend analytics and benchmarking really do surface consolidation and off-contract leakage you could not see across fragmented systems. A single intake front door really does reduce maverick requests and steer buyers toward the right catalog and the right paper. A contract repository really does put the negotiated terms in one place. None of that is theater. It is the work of finding money, and the good platforms do it well.

But a procurement organization lives and dies by two numbers, not one. There is the negotiated saving — the value identified at the table, on the sourcing award, in the consolidation analysis. And there is the realized saving — the value that actually reaches the P&L after every requisition, PO, and invoice has committed for the year. The category's own literature concedes, directionally, that a meaningful share of negotiated savings never survives the trip to the second number. That gap has a name on your side of the house: leakage. And leakage is not a discovery problem. You already discovered the saving. It is an enforcement problem.

Watch the verbs the spend-management and sourcing platforms actually use. They identify the opportunity. They negotiate the term. They recommend the preferred supplier. They route the requisition for approval. They flag the off-contract line and report the on-contract-spend percentage on a dashboard. Every one of those verbs happens around the transaction — before it as advice, after it as measurement — and the one moment that decides whether the saving is realized, the moment the PO or the invoice is authoritatively posted, happens in a system of record the analytics layer is reading from, not writing through.

This is why "on-contract spend" as a KPI is such a revealing metric. It is a scoreboard read after the game is over. A number that tells you what fraction of spend honored the contract is, by construction, a measurement of how much already did not. When it dips, you convene a meeting, trace the off-contract POs, and coach the buyers — all detective work, all after the money moved. The control that would have stopped the off-contract PO from committing lives somewhere else entirely: in an ERP configuration, a buyer's discipline, a policy in a PDF. The analytics narrate the leak with beautiful precision. They have no hand on the valve.

Concede the strongest version of the counter-argument, because it is real. The procurement-orchestration and intake-to-pay layer improves on the bolt-on dashboard: it puts a single front door in front of the mess, and it coordinates routing across the ERP and the source-to-pay suites so a request moves through the right sequence of systems. That genuinely cuts maverick requests and shortens cycles. But orchestrating and reporting across systems the layer does not own is not the same as the transaction executing on a fabric where the control is enforced at the moment of action. A conductor who can route the request to the ERP, and read back what the ERP decided, is still not the hand that posts — or refuses to post — the PO. The authoritative act, and therefore the enforcement, happens in a runtime the orchestration layer sits on top of.

Entroid moves the control from beside the runtime into the runtime. On the Composable Process Fabric, procurement is not a suite of intake forms, approval routing, and spend dashboards parked next to the ERP. It is modelled, executed, and governed as one workflow — intake to sourcing to requisition to PO to receipt to invoice to pay — composed from five primitives on a shared ontology, in a single runtime. The negotiated contract price, the budget envelope, the segregation-of-duties rule, and the authority threshold are not advice printed on a screen. They are carried inline by a Deterministic Workflow: the fixed, rule-governed step every action must clear to proceed. And a Connector is the only primitive that can write to the ledger, under authentication, authorization, and audit.

So the enforcement point moves to where it was always supposed to be — the moment of action. An off-contract line does not route to a dashboard for later scrutiny; it has no path to the posting step at all, because the workflow evaluates the requested price and supplier against the governing contract in the ontology before a Connector is ever asked to write. An over-threshold PO does not commit and then trip an alert; the authority gate refuses to route it. The requisition that would breach the budget envelope is not reconciled at month-end; it is non-executable until the envelope or the authority changes. The realized saving stops being a number you measure after the fact and becomes a property of what the fabric allowed to happen.

• Contract price & terms — the workflow checks the requested price and supplier against the live contract inline; an off-contract or off-price line has no route to the posting Connector.
• Budget envelope — the threshold is evaluated at the moment of commit, so the requisition that would breach it is refused, not reconciled later.
• Authority limits & SoD — the over-threshold PO cannot route past the authority gate, and raise-and-release cannot resolve to one identity.
• Evidence as byproduct — because the control is the executing step, every refusal and every passing action drops an immutable, per-action record; the audit trail is the exhaust of running, not a quarterly collection project.

Be exact about what this claim is. Consider — illustratively — a buyer who raises a PO to a supplier at a price above the negotiated tier, or outside contract entirely. The workflow resolves the line against the governing agreement in the ontology and finds the mismatch before any write is attempted; the payment instruction is simply never emitted at the non-compliant price. There is no leaked transaction to flag next month, because there is no posted transaction to find. That is an architectural property of the design — the fabric is built so a non-compliant action has no execution path — not a measured outcome from a named deployment, and not a claim that nothing else in the market can block anything. Narrow pre-execution stops exist elsewhere. The distinction is that here, inline deterministic enforcement is a native property of the fabric across the whole procurement process, carrying a provable per-action record with it, rather than a check bolted onto one runtime for one class of spend.

Do not let the argument tip into over-claiming, because that is the fastest way to lose a CPO who has actually run sourcing. Enforcement does not find the saving. Negotiation finds it. A competitive event finds it. Consolidation analysis finds it. The fabric cannot invent a better price; it can only refuse to let anyone spend past the price someone already fought for. The identification work — sourcing strategy, benchmarking, category management — remains indispensable, and the fabric is worth nothing without it. What changes is that the value discovered upstream is no longer left to survive on discipline and dashboards. It is held in place by the runtime that executes the buy.

And this is not a fantasy of zero integration. The fabric runs over your existing estate through governed Connectors; it does not demand you rip out the ERP or the rails you already run. Plenty of spend will keep flowing through systems the fabric does not yet govern — a legacy portal, a one-off payment pushed through a spreadsheet and a phone call. On that surface, detective spend analytics remains exactly the right layer, and the honest posture is to keep investing in it there. The point is not to retire detection on day one. It is to shrink the surface that needs it — process by process, Connector by Connector — until "we identified the saving" and "we realized the saving" stop being two different numbers with a gap between them.