The accountable but powerless service integrator at Haverstock
Haverstock is a public-sector agency with 4,500 staff and four outsourced providers: data centre and hosting, application support, wide-area network, and field engineering. An internal SIAM team of nine people acts as service integrator and is accountable in its charter for end-to-end incident, problem and change management. In practice, the providers' contracts are managed by a central procurement team, and provider account directors routinely escalate to the CIO's office without involving the SIAM team. In the last quarter, 38% of priority-1 incidents bypassed the SIAM team, mean time to restore was 11 hours against a 5-hour target, change success was 82%, and the change advisory board is chaired by the hosting provider. There is no shared priority matrix: the network provider treats a site outage as priority 3 while the agency treats it as priority 1. The CIO has asked for a diagnosis before the next contract review.
1. Diagnose the root causes of Haverstock's poor end-to-end performance using SIAM governance principles. Distinguish symptoms from causes.
Read the worked answer
The symptoms are 11-hour restoration, 38% bypass and 82% change success. The causes are governance failures, not tool or effort failures.
First, accountability has been assigned without authority. The SIAM team is accountable for end-to-end processes, but procurement owns contracts and the CIO's office accepts escalations directly. The integrator cannot set priorities, enforce data sharing or hold providers to account.
Second, the escalation design rewards bypass. A provider that escalates to the CIO gets attention; one that follows the SIAM route may wait. That is a rational response to unclear decision rights.
Third, there is no shared priority model. The network provider's priority 3 versus the agency's priority 1 is not a disagreement about severity; it is the absence of a single end-to-end impact assessment owned by the integrator.
Fourth, the change advisory board chair is the hosting provider, which is both a participant and a decision-maker. That creates a conflict of interest and weakens cross-provider change assessment.
The causes therefore sit in decision rights, escalation behaviour, priority ownership and forum design. Training the SIAM team or buying a new tool would not fix them.
2. Propose a decision-rights and escalation model that would give the service integrator authority proportionate to its accountability. Include the key RACI changes.
Read the worked answer
The SIAM team should own the end-to-end process decisions, while providers own delivery within their towers. Concretely:
- Priority-1 declaration and major incident control: SIAM incident manager decides, using an agreed business impact matrix. Providers are consulted but do not set the final priority.
- Escalation: a single path from provider operational contact to SIAM duty manager, then to the SIAM board, then the CIO. CIO office should redirect provider escalations back through SIAM unless it is a SIAM board escalation.
- Change: SIAM change owner chairs the CAB or delegates to an independent chair; the hosting provider becomes a CAB member, not chair. Emergency change authority sits with SIAM for cross-provider changes.
- Contract performance: procurement retains commercial ownership, but SIAM supplies the performance evidence and has a veto on service credits and renewal recommendations.
- Problem management: SIAM owns cross-provider problem records and can assign actions across providers.
In RACI terms, SIAM is accountable for end-to-end incident, problem and change; providers are responsible for their tower actions; procurement is consulted on contractual changes; the CIO is informed and acts as final escalation. The contracts must reference these decision rights, or the model remains informal.
3. The CIO asks whether to issue a directive telling providers to stop bypassing the SIAM team. How would you advise, and what would you do instead?
Read the worked answer
A directive alone would be weak advice. It addresses the visible behaviour, not the incentive. Providers bypass because the CIO's office responds and because the SIAM team cannot resolve their issues quickly. A directive may reduce the 38% figure temporarily while pushing escalation underground.
Instead, I would combine three actions. First, publish the escalation protocol and have the CIO visibly redirect the next few bypass attempts back to SIAM, with a service-level commitment that SIAM will respond within a stated time. Second, fix the SIAM team's capacity and decision rights so the redirect is credible: duty manager cover, a priority matrix and authority to convene cross-provider bridges. Third, measure the bypass rate and report it to the SIAM board, not as a compliance statistic but as an indicator of whether the protocol works.
The CIO should also address the procurement disconnect. If procurement continues to manage provider performance without SIAM evidence, providers will keep choosing the route that protects their contract position. The directive should therefore be backed by a contract change notice and a revised escalation clause at the next review.