Skip to content

FinOps Certified Professional (FOCP-Pro) - Practice Questions

15 questions at professional level: running a FinOps practice, not just performing its tasks. Covers operating model, maturity, cross-functional influence, and scaled governance.

Cert page: exams/finops/certified-professional/

Question 1

Scenario: A new FinOps practice is being stood up. What comes first?

A. Buy a cost management tool B. Establish allocation and visibility, since nobody can act on cost they cannot see or does not own C. Negotiate commitments D. Set reduction targets

Answer **Correct: B** **Why:** The Inform phase is first for a reason: targets set before allocation is trustworthy produce arguments about the numbers rather than action. Buying tooling before deciding the allocation model usually means configuring the tool twice.

Question 2

Scenario: Which describes the FinOps maturity model?

A. A certification level B. Crawl, Walk, Run applied per capability, so an organization is at different levels for different capabilities simultaneously C. A single organizational score D. A vendor assessment

Answer **Correct: B** **Why:** Maturity is per capability, which matters practically: a team can be at Run for allocation and Crawl for forecasting. Chasing Run everywhere at once is how practices stall; picking the capability whose immaturity is currently costing most is how they progress.

Question 3

Scenario: Engineering treats FinOps as the cost police.

A. Enforce harder B. Reframe the function as enablement: supply data and tooling engineers want, celebrate efficiency wins publicly, and make the practice about business value rather than cost minimization C. Escalate to leadership D. Set mandatory quotas

Answer **Correct: B** **Why:** Adoption is the binding constraint on a FinOps practice, and enforcement without relationship produces compliance theatre: tags applied meaninglessly, recommendations dismissed. The principle that FinOps optimizes for business value, sometimes by spending more, is the reframing that lands.

Question 4

Scenario: A commitment portfolio must be managed across a large estate.

A. Each team buys its own B. Centralize commitment management so it can be pooled across teams, with a laddered expiry schedule and utilization and coverage tracked continuously C. Buy annually in one purchase D. Avoid commitments

Answer **Correct: B** **Why:** Pooling raises utilization because one team's shortfall offsets another's excess. Laddering expiries avoids a cliff where a large share of the portfolio renews at once, which is both a budget shock and a negotiation weakness.

Question 5

Scenario: Shared costs (support, networking, shared clusters, observability tooling) must be allocated.

A. Leave them unallocated B. Choose and document a method: proportional to direct spend, even split, or usage-based where measurable, and apply it consistently C. Charge them to the FinOps team D. Split by headcount always

Answer **Correct: B** **Why:** No method is objectively correct, so consistency and transparency matter more than precision. What breaks trust is changing the method quietly, which makes teams' month-over-month numbers move for reasons they cannot explain.

Question 6

Scenario: Unit economics must be established for a multi-product organization.

A. One organization-wide metric B. Per-product unit metrics tied to how each product creates value, rolled up where comparison is meaningful C. Cost per employee D. Total spend only

Answer **Correct: B** **Why:** A single metric across dissimilar products compares nothing useful. The value of unit economics is that it makes a rising bill defensible: cost per transaction falling while spend rises is a growth story rather than a waste story.

Question 7

Scenario: Cloud spend must be forecast for annual budgeting.

A. Extrapolate the trend B. Combine bottom-up team forecasts with top-down trend analysis, incorporate planned launches, migrations, and decommissions, and track accuracy to improve the method C. Use last year plus 10% D. Ask the provider

Answer **Correct: B** **Why:** Bottom-up captures planned events that trend analysis cannot see; top-down catches the optimism in team estimates. Reconciling the two is the professional-level skill, and measured accuracy is what makes the next cycle better.

Question 8

Scenario: A multicloud estate must be reported on consistently.

A. Separate reports per provider B. Normalize to a common data model, ideally the FOCUS specification, so allocation and unit metrics are comparable across providers C. Report only the largest provider D. Convert everything to instance hours

Answer **Correct: B** **Why:** Each provider's billing export uses different column names and semantics for the same concepts, so per-provider reports cannot be compared without a translation layer. FOCUS is that layer as an open specification rather than a bespoke internal one.

Question 9

Scenario: Governance must prevent waste without blocking delivery.

A. Manual approval for every resource B. Guardrails: policies restricting instance families and regions, mandatory tagging enforced at creation, budget alerts, and automated cleanup of untagged or expired resources C. No controls D. Quarterly audits only

Answer **Correct: B** **Why:** Guardrails preserve self-service within bounds, which is the model that survives contact with engineering. Manual approval moves the bottleneck to a person and gets routed around; quarterly audits find the waste after months of paying for it.

Question 10

Scenario: Leadership asks for a 20% cost reduction target.

A. Apply it uniformly across teams B. Analyze where waste and opportunity actually sit, propose a plan with specific initiatives and owners, and surface where the target would harm business value C. Accept and cascade D. Refuse

Answer **Correct: B** **Why:** A uniform cut punishes efficient teams and rewards wasteful ones, since the efficient team has nothing easy left to give. Making the trade-off visible where the target conflicts with growth is the practitioner's job, not resistance to it.

Question 11

Scenario: Sustainability reporting is added to the FinOps remit.

A. Unrelated to FinOps B. Closely related: the same usage data underpins both, and many efficiency actions reduce carbon and cost together, though the two objectives occasionally diverge C. Replaces cost reporting D. Only relevant to hardware

Answer **Correct: B** **Why:** Rightsizing, decommissioning, and workload scheduling improve both. Divergence appears in regional placement, where the cheapest region and the lowest-carbon region are frequently different, which is a business trade-off rather than an optimization.

Question 12

Scenario: A FinOps practice must scale beyond a central team.

A. Hire more central practitioners B. Embed champions in engineering teams, provide self-service data and tooling, and keep the central function focused on rate negotiation, standards, and enablement C. Centralize all decisions D. Outsource it

Answer **Correct: B** **Why:** A central team scaling linearly with the estate becomes the bottleneck. The federated model puts decisions with the teams that make the technical choices, while centralizing the things that genuinely benefit from scale: contracts, tooling, and standards.

Question 13

Scenario: Which metric best indicates the FinOps practice itself is working?

A. Total spend reduction B. A portfolio: unit cost trend, allocation coverage, commitment utilization and coverage, forecast accuracy, and time to act on anomalies C. Number of recommendations generated D. Tool count

Answer **Correct: B** **Why:** Total spend can fall for reasons unrelated to the practice, or rise while efficiency improves. Recommendation counts measure activity rather than outcome. Time to act is the one that reveals whether insight is converting into change.

Question 14

Scenario: A large migration is planned and finance wants cost certainty.

A. Promise a fixed number B. Provide a modeled range with stated assumptions, identify the drivers of variance, and commit to tracking actuals against the model with regular re-forecasts C. Refuse to estimate D. Use the on-premises cost as the estimate

Answer **Correct: B** **Why:** A range with named assumptions is honest and actionable; a single number is a hostage. Using the on-premises cost as the estimate is the specific trap, because lifting and shifting an over-provisioned estate reliably costs more, not less, until it is rightsized.

Question 15

Scenario: Two teams dispute how a shared platform's cost is allocated to them.

A. Split it evenly and move on B. Return to the documented method, check whether the underlying usage data supports a more accurate split, and if the method is genuinely wrong, change it transparently and restate prior periods C. Escalate to leadership D. Leave it unallocated

Answer **Correct: B** **Why:** Allocation disputes are usually data disputes in disguise. Restating prior periods when the method changes is what preserves trust; changing it silently makes every team's historical trend meaningless and invites the next dispute.

Where to go deeper