Blanket POs for services (consulting, professional services, contingent labor, marketing services, legal advisory) are common but problematic. The instrument was designed for goods, where quantities and unit prices can be verified at receipt. Services lack that verifiability at the same point in the workflow.
Many companies use blanket POs for services in the same way they use them for goods, with predictable failure modes. Scope creeps beyond what the blanket authorized. Rates drift above the agreed schedule. Invoices arrive without clear linkage to deliverables. The matching workflow fails systematically.
Services blankets can work, but only with a different approval model that accounts for what services actually look like as they are consumed.
Why Services Are Different
Three structural differences from goods blankets matter.
No physical receipt event
Goods get delivered, received, and inspected at a discrete moment. Services accumulate over time. The buyer cannot inspect 80 hours of consulting work at the moment those hours were performed. Verification happens through time tracking, deliverable acceptance, or both, on a different cadence from the work itself.
Variable consumption rate
Goods consumption is bounded by quantity. A blanket for 1,000 widgets at $50 each is exhausted when 1,000 widgets have been ordered. Services consumption depends on time spent, which can vary substantially based on project scope and resource utilization. The same engagement might consume 500 hours or 1,500 hours depending on what gets uncovered.
Rates can vary by resource
A services blanket might cover work by multiple resource levels: principal, senior consultant, consultant, analyst. Each level has its own rate. The blanket cannot just track total dollars; it needs to track which rates apply to which work and ensure each invoice charges the right rate for the right resource.
The Three Common Service Blanket Models
Most service blankets fall into one of three patterns. Each one has different control requirements.
Hours based blanket
The blanket authorizes a defined number of hours at defined rate cards. Each release specifies the work and the expected hours. Invoices reference actual hours worked, which are matched against the release. Control point: time records.
Common for IT services, advisory work, technical contractors. Works well when the work can be reasonably scoped in hours and the rate card structure is stable.
Milestone or deliverable based blanket
The blanket authorizes a series of milestones or deliverables, each with a defined payment. Releases reference specific milestones. Invoices arrive when milestones are completed and accepted. Control point: deliverable acceptance.
Common for marketing services, software development projects, design work. Works well when the work can be decomposed into discrete deliverables with clear acceptance criteria.
Retainer or fixed monthly fee blanket
The blanket authorizes a defined monthly fee for ongoing services, often with a defined scope of activity per month. Releases (if any) are monthly. Invoices arrive monthly at the agreed amount. Control point: scope adherence within the monthly engagement.
Common for ongoing professional services (PR, marketing, IT support). Works well when the service is genuinely ongoing rather than project specific.
Approval at the Right Point
For services blankets, approval timing matters. Approving only at the blanket creation, with releases routing through lightweight workflows, leaves significant scope drift and rate creep unmonitored.
Approval at the SOW or work order level
Each release against a services blanket should be backed by a statement of work (SOW) or work order that specifies the scope, deliverables, timeline, and expected effort. The release approval is essentially approving the SOW, not just authorizing the dollar release.
Without the SOW level approval, the release approver has no basis for judging whether the work being authorized matches the original intent of the blanket.
Approval at the time tracking layer
For hours based blankets, weekly or biweekly time approval is the operational control. The work performed in the period gets reviewed against the SOW. Hours that exceed the SOW expectation or that fall outside the agreed scope get flagged before they accumulate into invoices.
This is the most labor intensive control but also the most effective. Without it, hours accumulate invisibly and the issues only surface at invoice time when changing direction is harder.
Approval at deliverable acceptance
For milestone based blankets, the acceptance step is the substantive control. The buyer reviews the deliverable against the agreed acceptance criteria, accepts or rejects, and the acceptance triggers the invoice approval.
Acceptance reviews that become routine rubber stamps defeat the control purpose. The acceptance step needs to be genuinely substantive.
Tracking Against Rate Cards and SOWs
Services blankets typically reference an underlying rate card or fee schedule. The matching workflow needs to verify that invoiced charges align with the rate card.
Three checks should run on each services invoice:
- Resource and rate match: the resource level invoiced (principal, senior, consultant) charges the rate specified in the rate card for that level. Variances flag for review.
- Hours match: hours invoiced match hours reported in time tracking systems. Discrepancies between invoiced and tracked hours indicate either time tracking errors or unauthorized work.
- Scope match: the work invoiced falls within the scope of the SOW or work order referenced. Out of scope work is a separate authorization question, not an automatic invoice approval.
The Risks Specific to Services Blankets
Four risks show up consistently in services blankets that goods blankets do not face in the same way.
Scope creep
Work expands beyond the original SOW without formal scope amendment. The hours or fees accumulate, the blanket limit gets consumed faster than expected, and the buyer is faced with either authorizing additional work after the fact or pushing back on already completed work.
Rate creep
The rate card on the original blanket was negotiated at certain levels. Over the validity period, the supplier escalates rates citing market changes, resource scarcity, or other justifications. If the blanket does not have clear rate change provisions, the buyer ends up paying the new rates on top of original work that was scoped at old rates.
Resource substitution
The original engagement was scoped to use senior consultants at $250 per hour. The supplier substitutes principal level resources at $400 per hour, citing scope changes or resource availability. Without resource and rate matching at the invoice level, the substitution can quietly inflate the engagement cost.
Time card padding
Hours invoiced exceed hours actually worked, either through aggressive rounding, inclusion of administrative time that should be supplier overhead, or outright padding. Without independent time validation, this is difficult to detect.
Service Blanket Governance
The governance discipline for services blankets has three elements that distinguish it from goods blankets.
- Every release backed by an explicit SOW with scope, deliverables, expected effort, and acceptance criteria
- Periodic time and progress review against the SOW, not just at invoice time but weekly or biweekly during active work
- Rate card lock at blanket creation, with explicit amendment process if rates need to change during the validity period
- Quarterly review of supplier resource composition: are the actual resources used aligned with what the SOW specified, or has substitution drifted the engagement up the rate card
These add overhead compared to goods blankets, but the overhead is proportional to the risk. Services spend that runs without these controls regularly produces unpleasant year end surprises.
Start Here
Pull a sample of services blanket releases from the last quarter and walk each one through the controls: was there a defined SOW, was scope tracked, were hours validated against time records, were rates validated against the rate card. The gaps are the diagnostic.
If most releases lack one or more of these controls, the services blanket discipline needs investment. The risk profile of services spend is materially higher than goods spend, and the control level should reflect that.
