Choosing a service integrator model at Meridian Freight
Meridian Freight is a European logistics operator with 9,000 staff and revenue of €2.1bn. IT is delivered by three providers: a data centre and hosting provider (€18m/year), a network provider (€6m/year) and an applications support provider (€11m/year). Each provider runs its own service desk, its own incident tool and its own SLA report. The retained IT team of 30 people manages contracts and architecture but runs no operational process. Major incidents are coordinated ad hoc by whichever provider escalates loudest; in the last quarter three P1 outages took between 4 and 9 hours to restore, and the post-incident reports contradicted each other. The board has approved a 'SIAM programme' but has not decided whether to build an internal service integrator, appoint one of the existing providers as integrator, or hire a specialist SIAM provider. The CIO favours appointing the hosting provider because 'they already know our estate'.
1. Evaluate the three sourcing options for the service integrator role and recommend one, showing the criteria you would use.
Read the worked answer
Compare the options against explicit criteria before recommending: impartiality, capability, time to value, cost structure, control, reversibility and exit.
- Internal integrator. Highest control and impartiality, and accountability stays with Meridian. But the 30-person retained team does contracts and architecture, not operational process, so building incident, problem, change and service level capability will take roughly 12–18 months and a large fixed cost. Suits organisations where SIAM is strategic and the provider estate is stable.
- Lead provider as integrator. Fastest and cheapest on paper because the hosting provider already has tooling and estate knowledge. The fatal flaw is conflict of interest: the integrator must arbitrate when the hosting provider's own service causes a P1, and must publish honest performance data about itself. The three contradictory P1 reports show that impartial arbitration is precisely what Meridian lacks.
- Specialist SIAM provider. Brings process and tooling capability quickly and is contractually impartial, but adds a layer, creates a new dependency and requires strong retained governance so accountability is not simply outsourced again.
Recommendation: a hybrid - a small internal core accountable for end-to-end service, governance and provider performance, augmented by a specialist for process design, tooling integration and transition support for 12–18 months, with an option to insource. This protects impartiality, builds capability and can be tested against contract exit dates and the business cost of the three P1 outages.
2. The CIO argues that the hosting provider should be integrator because 'they already know our estate'. What counter-arguments and safeguards would you put to the board?
Read the worked answer
Knowledge of the estate is a real advantage, but it is not the same as fitness for the role. The service integrator must be able to say that a provider is failing, enforce a single major incident process, and publish performance data that may embarrass a sister business unit. A provider that also delivers a large part of the estate has a commercial interest in attributing faults elsewhere, and the other two providers will reasonably suspect that their data and escalations are being filtered.
Counter-arguments to present:
- Conflict of interest is structural, not personal. It cannot be fixed by goodwill; it must be designed out through disclosure, audit and separation of duties.
- Precedent from the incidents. Contradictory post-incident reports already show that providers will protect their position when no neutral party owns the end-to-end view.
- Market and exit risk. Appointing a lead provider makes the integrator hard to replace without destabilising the hosting service.
If the board still proceeds, safeguards should include: contractual obligations to disclose the provider's own performance and root-cause data; an independent assurance or audit right; a ring-fenced integration unit with separate reporting to the CIO; service credits tied to integration behaviour, not just uptime; and a defined exit and re-tender route. Without those, the model should not be approved.
3. Explain what SIAM is intended to add here that contract management and SLA reporting do not already provide.
Read the worked answer
Contract management manages each supplier relationship; SLA reporting measures each supplier against its own commitments. Neither produces an end-to-end service. Meridian's symptoms - three P1s with 4–9 hour restoration, ad hoc coordination, contradictory reports - are integration failures, not supplier failures. Each provider may be meeting its own SLA while the customer experience is still poor.
SIAM adds four things. First, single point of accountability for end-to-end service performance across providers, so someone owns the outcome rather than the inputs. Second, integrated processes: one priority matrix, one major incident process, one escalation route and cross-provider problem management, so handoffs do not become blind spots. Third, a single service view: shared data, a common tool or federated tooling, and reporting that shows the customer journey rather than three separate scorecards. Fourth, governance and arbitration: a forum and a documented mechanism that settles disputes between providers on evidence rather than volume.
The distinction to make to the board is between orchestration and integration. Orchestration coordinates tasks; integration owns the service outcome and the process model. Meridian's problem is the latter, which is why buying better contract dashboards would not fix it.