By Jim Parke |
A major programme can have a signed contract for every work package and still struggle to deliver one working system. The civil contractor finishes a structure; the technology supplier needs access to test its equipment; the future operator needs a design change to maintain it safely. Each organisation can meet its own obligations while the interfaces between them remain unresolved.
More contract detail may appear to be the answer. So may an appeal to everyone to collaborate. Neither instruction is sufficient on its own. A contract can specify responsibilities and routes for resolving disputes. Trust helps people share information early enough for those routes to be useful. A sponsor needs to design for both.
Separate cooperation from coordination
Organisational researchers Ranjay Gulati, Franz Wohlgezogen and Pavel Zhelyazkov distinguish cooperation from coordination. Cooperation concerns whether organisations are willing to work towards a common result. Coordination concerns how their different activities fit together in time and space. Teams can sincerely support a programme and still miss an interface because their schedules, information systems or assumptions do not match. Conversely, a detailed sequence of obligations will not produce candid early warnings if participants fear that disclosing a problem will be used against them.
That distinction changes the sponsor's questions. Instead of asking only whether every party has a scope of work, ask what happens when the scopes touch. Who owns the space between two packages? When must a supplier share a design assumption? Which organisation can resolve a conflict before it reaches the critical path? What will each participant gain, or lose, by surfacing a risk early?
The organisational architecture matters as much as the diagram of contracts. Research by Juliano Denicol, Andrew Davies and Stephen Pryke describes megaproject organisation across internal, inter-organisational and wider system levels. Their study, which included Crossrail among six cases, emphasises the client's role in configuring the delivery organisation. The useful practical point is to examine relationships across the whole delivery system, including the sponsor and future operator, before treating an individual contract as a complete unit of design.
Write contracts that support useful behaviour
Contracts matter. They allocate work, price, risk and authority; set procedures for change; and give parties a basis for acting when expectations differ. Clear obligations can make collaboration less hazardous because the parties know the limits of their commitment. Research on inter-organisational relationships by Reinhard Bachmann and Andrew Inkpen examines how institutional arrangements, including legal rules, industry standards and professional norms, can support the development of trust.
The question is what behaviour the agreement makes attractive. If each contractor is rewarded for protecting its own milestone while programme-wide testing is somebody else's problem, the sponsor should expect difficult interfaces. If notification of a potential delay is treated only as the first move in a claim, important information may arrive late. The sponsor needs to examine incentives and processes across packages, not just whether each package is enforceable.
None of this requires asking a contractor to absorb unlimited risk for the sake of partnership. A workable agreement should say who can make an adjustment, what information must be shared, how its cost will be assessed and when an unresolved issue goes to a joint forum. It should also distinguish a genuine programme problem from a request to shift one party's agreed risk to another.
Evidence does not support a simple opposition between formal agreements and relationships. In a meta-analysis covering 33,051 inter-organisational relationships across 149 studies, Zhi Cao and Fabrice Lumineau found contractual governance positively related to trust and relational norms. They also found that contracts, trust and relational norms jointly improved satisfaction and relationship performance and reduced opportunism. Their evidence concerns inter-organisational relationships generally, rather than megaprojects specifically. It does not mean every additional clause builds trust; it shows why formal and relational arrangements should be designed together.
Give the interfaces somewhere to be resolved
An interface register can show dependencies, but a register cannot resolve them. A programme should identify the people who can make a decision across packages and give them a regular forum with access to the relevant designs, schedules and risks. Where a matter exceeds their authority, the route to the sponsor should be short and explicit.
Shared information is part of that arrangement. A common model or digital platform can expose conflicting assumptions, provided each party knows what information is authoritative and when it must be updated. More data does not help when teams use different definitions of completion or cannot see the effect of a proposed change on the operator. The meeting, the data and the decision rights have to work together.
Crossrail's published learning legacy offers a concrete reminder of the difficulty. Its systems integration material describes the challenge of assuring a railway design delivered through many separate contracts. That experience does not supply a universal contract template. It illustrates why the client must retain responsibility for integration of the system rather than assume that compatible parts will emerge automatically from individually managed packages.
Build trust through repeated, observable conduct
Trust is sometimes presented as a sentiment. In a programme, it is tested in ordinary interactions. Do parties disclose a foreseeable problem before it becomes a formal claim? Do they explain why a request has been refused? Does the sponsor apply a change process consistently when the result is inconvenient? Can a technical team correct an assumption without being punished for raising it?
Joint workshops and a shared purpose can help, but conduct over time is more persuasive. The sponsor should create opportunities for teams to solve modest problems together before a major dispute arrives. It should also be ready to enforce the agreed process fairly. Trust is weakened if cooperation means surrendering a legitimate right, or if formal rights are used to avoid discussing the substance of a shared problem.
Parties will continue to have different interests. A delivery contractor, public sponsor and operator will not value every outcome in the same way. Good governance allows those differences to be expressed early and dealt with through a process that all can understand. It does not require pretending the differences have disappeared.
Test the whole system before declaring success
Programme reporting often follows contract boundaries: package complete, payment approved, milestone achieved. The sponsor also needs evidence that the combined system can be commissioned, operated and maintained as intended. Crossrail organised its integration work around four collective tests: safe, operable, maintainable and performing as intended. That is a more demanding question than whether each contract has delivered its part. It calls for shared tests, named owners of unresolved interfaces and enough time to correct what testing reveals.
The structure awaiting equipment and an operator's design change makes the test concrete. The immediate question is not whether the parties have been sufficiently friendly or whether the contract is sufficiently thick. It is whether the programme established a timely way to expose the dependency, decide how to change the work and allocate its consequences. Contracts set the ground rules. Trust makes it more likely that people will use them while there is still time to act.
Further reading
- Ranjay Gulati, Franz Wohlgezogen and Pavel Zhelyazkov, The Two Facets of Collaboration: Cooperation and Coordination in Strategic Alliances (2012).
- Reinhard Bachmann and Andrew C. Inkpen, Understanding Institutional-based Trust Building Processes in Inter-organizational Relationships (2011).
- Zhi Cao and Fabrice Lumineau, Revisiting the interplay between contractual and relational governance (2015).
- Juliano Denicol, Andrew Davies and Stephen Pryke, The organisational architecture of megaprojects (2021).
- Crossrail Learning Legacy: Systems Integration and Technical Assurance.