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. Negotiate commitments with the cloud providers before usage is understood B. Establish allocation and visibility; nobody can act on unseen cost C. Set cost reduction targets for every team before the allocation model is agreed D. Buy a cost management tool and connect every account, so dashboards exist before any allocation model is argued over

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 vendor assessment B. Crawl, Walk, Run applied per capability, so an organization sits at different levels for different capabilities C. The FinOps lifecycle of Inform, Optimize and Operate, which every team completes once, in that order, and is then considered fully mature D. A certification level

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, by adding automated shutdowns and a weekly report naming every team that exceeded its tagged budget B. Escalate to leadership and ask executives to require that every team cut spend by a fixed percentage each quarter, with exceptions approved by finance C. Reframe it as enablement: give engineers the data and tooling they want and celebrate wins D. Set mandatory spending quotas for each team and block any deployment that would take the team over its quota this month

Answer **Correct: C** **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. Avoid commitments entirely and stay on on-demand pricing for flexibility B. Buy annually in one purchase at the start of the fiscal year to get the best discount C. Centralize commitment management so it can be pooled across teams, with laddered expiries and continuous utilization tracking D. Each team buys its own to match its budget

Answer **Correct: C** **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. Charge them to the FinOps team C. Split by headcount in every case, since headcount is easy to audit and every team agrees on its own team size D. Choose and document a method, such as proportional to direct spend, and apply it consistently

Answer **Correct: D** **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. Per-product unit metrics tied to how each product creates value B. Total monthly spend per product, compared across all products on the same scale C. Cloud cost per employee, reported for the whole organization every quarter D. One organization-wide metric such as total cost divided by total revenue, applied to every product

Answer **Correct: A** **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. Ask the provider B. Combine bottom-up team forecasts with top-down trend analysis, include launches and migrations, track accuracy C. Use last year's spend plus a flat 10% for growth, then lock the number for the whole year and report any overrun to finance as a variance D. Extrapolate the trend from the last six months of billing data and add a fixed buffer for growth and expected price changes

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. Report only the largest provider C. Convert every provider's usage to normalized instance hours, since compute is the one unit all providers bill in a comparable way D. Normalize to a common data model, ideally the FOCUS specification, so unit metrics are comparable across providers

Answer **Correct: D** **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 new resource by the central cloud team, with finance reviewing each request against the quarterly budget before launch B. Guardrails: policy limits, mandatory tagging at creation, budget alerts, and automated cleanup C. Quarterly audits only D. No controls, trusting engineers to clean up what they launch and finance to spot anything unusual on the invoice

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. Refuse B. Accept the target and ask each team to pledge its own saving, with the pledges totalled and reported monthly C. Apply it uniformly across teams, so every team cuts 20% of its own spend and the target is met fairly in every cost center, with progress reported monthly to the executive sponsor D. Analyze where waste sits, propose owned initiatives, and flag where the target harms value

Answer **Correct: D** **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 carbon and cost reporting, and many efficiency actions cut them together, though they sometimes diverge C. Replaces cost reporting D. Only relevant to hardware, because carbon emissions come from the manufacture and shipping of physical servers rather than from how cloud workloads are scheduled and sized

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. Embed champions in engineering teams, provide self-service data and tooling, and keep the central function on negotiation and standards B. Centralize all decisions in the FinOps team so every purchase and design choice gets one consistent review C. Outsource it D. Hire more central practitioners until each one covers about ten engineering teams

Answer **Correct: A** **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. A portfolio: unit cost trend, allocation coverage, commitment utilization, forecast accuracy, time to act B. Number of optimization recommendations generated by the tooling each month, plus the count of dashboards published to teams C. Tool count D. Total spend reduction year on year, measured across the invoices of all accounts and business units, with a target set by finance

Answer **Correct: A** **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. Use the on-premises cost as the estimate, adjusted by the vendor's published list-price discount for the chosen instance types and applied uniformly to every workload in the migration plan B. Promise one fixed number to finance now and hold every team to it for the whole migration, whatever the migration reveals C. Provide a modeled range with stated assumptions, variance drivers, and regular re-forecasts against actuals D. Refuse to give any estimate until the migration is finished and actual costs have been measured for a full quarter

Answer **Correct: C** **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. Return to the documented method, check whether usage data supports a better split, and if it is wrong, change it transparently and restate prior periods B. Escalate to leadership and let the executive sponsor decide the split C. Leave it unallocated D. Split it evenly and move on, noting the decision in the next quarterly review

Answer **Correct: A** **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