SIAM operation, performance management and continual improvement
Domain 7 tests how the running model is operated, measured and improved - including the metrics that reveal whether integration is actually happening.
Operating the ecosystem
Day-to-day operation means running the integration processes: service desk and major incident, change and release coordination, cross-provider problem management, request fulfilment, capacity, availability and continuity, plus relationship, contract and tooling upkeep. Cadence makes it work: a daily operational call for exceptions and changes, a weekly service review for performance and incidents, a monthly service performance review with providers, and a quarterly business review with service owners and the governance board.
Performance management
- Build a measurement framework from critical success factors down to a small set of KPIs; measure the end-to-end service as well as each provider.
- Keep one version of the truth: a single reporting source with views tailored to business, integrator and provider audiences.
- Watch interfaces, not just towers - many failures occur where providers hand off.
- Track trend, risk and cost alongside availability; a balanced scorecard stops the model optimising one dimension.
- Guard against gaming: metrics a provider can influence alone may be met while the service still disappoints.
Continual improvement
Improvement is a SIAM process, not an afterthought. Inputs include performance reports, incident and problem trends, audit findings, customer feedback and maturity assessments; outputs are prioritised, owned actions in an improvement register with defined value and verified benefit. The integrator orchestrates improvement across providers, every provider contributes, and the governance board funds what matters.
Worked example
End-to-end availability runs at 99.2% against a 99.9% target, yet every provider meets its own SLA. Reasoning: the shortfall sits at the interfaces - duplicated monitoring, unaligned change windows, unclear dependency mapping. Action: joint dependency mapping in the integrated CMDB, a shared change calendar, and one end-to-end availability KPI reported by the integrator and reviewed monthly with the board until the trend reverses.
Traps that cost marks
- Reporting only provider SLAs and calling it end-to-end performance.
- Too many KPIs, or KPIs nobody acts on.
- Improvement actions with no named owner or benefit verification.
- Letting the SIAM model go stale as providers and services change.
- Assuming continual improvement belongs to providers alone.