Question 1

Who should own a steering project: finance or IT?

Accepted Answer

Neither one alone. A steering project is a business project owned by finance (or the relevant function), with an executive sponsor, and IT as an active partner on architecture, data flows, security and sustainability. The classic trap is a '100% IT' project that delivers a tool without adoption, or a '100% finance' project that creates unmaintainable technical debt. Good governance is tripartite: business, finance, IT.

Question 2

How do you align finance, business and data around a shared vision?

Accepted Answer

Start by sharing a common framework: decision processes, key indicators, responsibilities, language (reference data, definitions). Then equip that vision with a shared data model (xP&A) rather than application silos. Finally, set up transverse steering rituals (performance reviews, data committees, planning sprints) that force functions to confront their assumptions on the same numbers.

Question 3

Why is data quality critical for steering?

Accepted Answer

No steering tool, however sophisticated, can compensate for inaccurate, incomplete or inconsistent data. Poor-quality data produces biased analyses, erodes user trust, multiplies manual rework and ends up discrediting the function distributing it. Conversely, reliable and governed data accelerates decisions, reduces control loops and frees time for value-added analysis.

Question 4

What is reliable, governed data?

Accepted Answer

Reliable data is accurate, complete, up-to-date, traceable and consistent across systems. Governed data also has an identified owner (data owner), documented quality rules, a defined lifecycle (creation, validation, update, archiving) and quality KPIs being tracked. Governance is what ensures reliability holds over time, beyond the initial cleanup.

Question 5

Should data be structured before launching an EPM project?

Accepted Answer

Ideally yes — at least on the relevant scope. Launching an EPM on fragile source data exposes the project to spending its time fixing discrepancies rather than modelling value. In practice, both run in parallel: prioritised MDM/data quality streams are scoped upstream of the EPM perimeter, and governance is industrialised as the rollout progresses. It's one of the most underestimated success factors.