Two suppliers, one end-to-end failure
TransCargo is replacing its customs clearance platform. It has contracted a prime systems integrator, Vela, on a fixed-price, output-based statement of work, and a specialist cloud provider, Nord, directly for platform hosting. The contract makes Vela responsible for 'application integration' and Nord responsible for 'platform availability'. Three integration defects have fallen between the two suppliers: Vela says the platform network configuration is Nord's; Nord says the application retry logic is Vela's. Monthly service reviews focus on SLA attainment, and both suppliers are meeting their individual SLAs while the end-to-end customs submission fails about 6 percent of the time. The project is three months late. Knowledge transfer is described in the contract as 'reasonable cooperation', and key Vela staff rotate every four months. The PM proposes withholding payment and adding penalty clauses. The CFO wants to know whether to bring integration back in-house.
1. Diagnose the multi-supplier integration and contract problems.
Read the worked answer
The core problem is that the contract manages outputs and individual availability, while the business needs an end-to-end outcome: reliable customs submission. Both suppliers can meet their own SLAs while the service fails because no one owns the interfaces. The boundary wording is too vague: 'application integration' and 'platform availability' do not say who configures network paths, who diagnoses cross-layer retry behaviour, who tests end-to-end, or who accepts residual risk at the boundary.
Fixed-price, output-based contracts can encourage each supplier to optimise its own scope and reject work at the interface. Monthly SLA reviews reinforce this by measuring what is easy to measure rather than what matters. The absence of a joint risk register and joint problem-solving forum means defects become commercial arguments. Knowledge transfer described as 'reasonable cooperation' is unenforceable in practice, and four-month staff rotation undermines continuity and tacit knowledge.
This is a service integration failure, not simply a supplier performance failure. TransCargo has retained two contracts but not the capability to integrate them. The 6 percent end-to-end failure is the true measure, and it is invisible in the supplier scorecards.
2. Evaluate the PM's proposal to withhold payment and add penalty clauses.
Read the worked answer
Withholding payment and adding penalties may be contractually available, but as a first response they are unlikely to fix the integration problem and may make it worse. The suppliers are already meeting their individual SLAs, so each will argue that the failure lies on the other side of the boundary. Penalties can sharpen that argument, encourage defensive evidence collection and reduce willingness to share diagnostic information. That slows root-cause resolution and increases the chance that defects remain hidden until go-live.
A better sequence is to establish the facts jointly: map the end-to-end transaction flow, define interface responsibilities, reproduce the failures, and assign diagnosis to a joint team rather than exchanging blame. Then align commercial mechanisms with the outcome. That could include an end-to-end SLA with shared consequences, a joint risk and issue register, agreed interface specifications, and a change to the statements of work so integration testing and knowledge transfer are explicit deliverables. Penalties can be retained as a backstop for clear non-performance, but they should not substitute for service integration.
The PM should also recognise that the project's own integration and acceptance responsibilities cannot be outsourced entirely. TransCargo needs an informed customer role that can arbitrate technical boundaries and test end-to-end outcomes.
3. What should TransCargo retain in-house, and what should it change in the supplier set-up?
Read the worked answer
TransCargo should retain service integration and architectural decision authority. That means an internal owner for the end-to-end customs service, capability to define and test interfaces, contract and performance management, and enough technical knowledge to challenge supplier diagnoses. It should retain knowledge management for critical operational knowledge, even if suppliers deliver training. Bringing everything in-house is not automatically the answer; the CFO should first establish whether the organisation can integrate and govern multiple suppliers. If not, a service integrator role or a small retained integration team may be more realistic than rebuilding the whole platform.
The supplier set-up should change from output silos to outcome alignment. Introduce an end-to-end service description and SLA covering customs submission success, latency and recovery. Define interface responsibilities in a responsibility matrix that covers configuration, monitoring, testing, incident diagnosis and change. Create a joint integration board and joint problem management process with shared data. Replace 'reasonable cooperation' with specific knowledge-transfer obligations, named roles, rotation handover requirements and acceptance criteria. Align incentives so both suppliers benefit when the end-to-end outcome improves, and make joint risks visible to TransCargo's steering group.
Finally, the PM should stop treating the issue as a payment lever. The commercial route should reinforce the operating model: one end-to-end owner, clear boundaries, joint governance and measures that reflect the customer outcome.