How Should CSDM Fidelity Be Balanced Against Governance Debt?

[eriksandber]

CSDM gives organizations a prescriptive foundation for modeling services and the supporting enterprise landscape, but there is an architectural tradeoff that I rarely see discussed:  at what point does increasing model fidelity cost more to govern than the traditional business value it produces?

For a large hybrid enterprise with acquisitions, legacy, applications, shared platforms, cloud services, outsourced services, and inconsistent ownership maturity, how do experienced architects determine the minimum level of Business Application, Application Services, Technical Service, Service Offering, and CI relationship fidelity necessary to support impact analysis, change, risk, portfolio planning, operations, and executive reporting?

What objective criteria would determine whether another relationship or layer in the CSDM is worth creating and governing?

And what measures would you use to demonstrate that CSDM maturity is improving business decisions rather than simply increasing CMDB completeness?

[Mathew Hillyard]

Hi @eriksandber 

I notice you've just joined the community and have posted quite a few questions using the same level and style.

There is no one size fits all to any of your questions. They depend on the use case, organisation or situation. One must adapt to all three and tailor any piece of work to the customer culture and pragmatically apply ServiceNow solutions that achieve their business objectives without incurring unnecessary cost, risk, or technical debt.

Value metrics for CSDM need to focus on an organisation's key objectives. The CSDM Data Foundations Dashboard can show you where you are on the journey but not show you what value you're getting from that journey. Start with the "why" first. For example, if reduced MTTR is essential, what needs to enable that metric? One possible answer is Event Management or similar monitoring as the entry point (or even Vulnerability Response to present a remediation before an issue is discovered, thus reducing MTTR to zero in theory), Incident Management has a clear understanding of incident source, categorisation, causal CI and impacted service identification, assignment, meaningful SLAs and OLAs supported by a near- or fully-automated heal process to restore service.

I hope this helps!
Mat