# The CFO Dashboard: From Data Overload to Decision Intelligence
When Brian Olsavsky, Amazon's CFO, was asked in a 2024 internal town hall how many metrics he reviewed before approving the quarterly close, his answer stunned the audience: eleven. Not eleven hundred. Eleven. For a company processing 1.6 million orders per hour across 21 currencies, the discipline is almost violent in its simplicity. Compare that to the typical Fortune 500 CFO, who according to a 2025 Gartner survey receives 52 distinct financial metrics weekly, opens fewer than 14 of them, and acts on perhaps 6. The gap between what gets reported and what gets *decided* is the single largest source of waste in modern finance functions.
This lesson is about closing that gap. Not by adding more dashboards, but by ruthlessly subtracting from them.
The neuroscience here is unambiguous. Research from MIT Sloan (Malone & Bernstein, updated 2024) shows that executive decision quality degrades sharply once working memory exceeds 7 ± 2 items—the Miller threshold first identified in 1956. Every additional KPIKPIKey Performance Indicator, a measurable value that shows how effectively you're achieving a specific objective, tracked over time against a target.Voir la définition complète → on a dashboard beyond that point does not add information; it adds noise that crowds out signal. McKinsey's 2025 *State of the CFO* report quantified the cost: finance leaders monitoring more than 20 KPIs were
This is not an argument for ignorance. It is an argument for hierarchy. Great CFO dashboards are pyramidal: a top tier of 8–12 decision metrics, a second tier of 30–50 diagnostic metrics accessed on drill-down, and an underlying data lakedata lakeA data lake is a centralized repository that stores large volumes of raw data in its native format, from structured tables to unstructured files, until needed.Voir la définition complète → of thousands of operational measures that exist for the FP&A team, not the CFO.
In our work with mid-cap European industrials post-CSRD implementation in 2024, we've identified three recurring failure modes:
1. The Accountant's Trap. Dashboards become a P&L recital—revenue, gross margingross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.Voir la définition complète →, EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète →, net income—with no leading indicators. By the time a metric like "EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → margin" moves, the underlying cause is 90 days old.
2. The IT Vendor Hangover. Tableau, Power BIBITechnologies and processes that turn raw data into actionable insights via reporting, dashboards and analysis, so teams can decide based on facts rather than intuition.Voir la définition complète →, and Anaplan implementations between 2020 and 2023 created the illusion that more visibility equals more control. Unilever's former CFO Graeme Pitkethly noted in a 2023 *FT* interview that his team had built 187 dashboards—of which 12 were viewed monthly.
3. The Compliance Creep. IFRS 16 lease metrics, CSRD double-materiality KPIs, and OECD Pillar Two effective tax rate disclosures have piled onto executive reports without anyone removing legacy items. The result: dashboards optimized for auditors, not decisions.
Kaplan and Norton's Balanced Scorecard (BSC), introduced in 1992, remains the most coherent architecture for executive KPIKPIKey Performance Indicator, a measurable value that shows how effectively you're achieving a specific objective, tracked over time against a target.Voir la définition complète → design—but in 2026 it needs adaptation. The original four perspectives (Financial, Customer, Internal Process, Learning & Growth) were designed for CEO-level oversight. For a CFO dashboard, we recommend a modified four-quadrant model:
That's 11 metrics. Anything more belongs on tier two.
Where the Balanced Scorecard provides *structure*, OKRs (Objectives and Key Results, popularized by Andy Grove at Intel and embedded at Google by John Doerr) provide *intention*. The CFO dashboard should explicitly link each KPIKPIKey Performance Indicator, a measurable value that shows how effectively you're achieving a specific objective, tracked over time against a target.Voir la définition complète → to a quarterly Key Result.
Example from Adobe's finance organization (disclosed by CFO Dan Durn at a 2024 J.P. Morgan conference): Adobe's FP&A team runs a quarterly KR cycle where every executive dashboard metric must answer the question, *"What decision does this enable in the next 90 days?"* Metrics that fail that test are demoted to tier two. Between FY2022 and FY2024, Adobe reduced its executive financial dashboard from 31 to 9 metrics, and finance cycle time (close-to-decision) dropped from 14 days to 6.
A.P. Møller-Maersk's CFO Patrick Jany inherited a dashboard problem in 2022. The shipping giant had built, during the pandemic freight boom, a real-time "control tower" with 240+ metrics spanning 130 country operations. When ocean freight rates collapsed 78% between Q3 2022 and Q2 2023 (Drewry WCI), the dashboard didn't help executives act faster—it paralyzed them.
Jany's team executed a six-month redesign anchored on a question Kaplan calls "the so-what test": *If this metric moved 20% tomorrow, would anyone make a different decision?* Roughly 80% of metrics failed.
The new Maersk executive dashboard, deployed in October 2023:
The result, disclosed in Maersk's 2024 capital markets day: management identified the Red Sea / Houthi disruption impact on Q1 2024 routing economics within 48 hours of December 2023 escalation, versus an internal estimate of 3 weeks under the old system. That speed translated to an estimated $340M in protected EBIT through pre-emptive rate adjustments.
Vérification des acquis
1. According to the lesson, how many financial metrics does Amazon CFO Brian Olsavsky review before approving the quarterly close?
2. Per the McKinsey 2025 State of the CFO report cited in the lesson, finance leaders tracking more than 20 KPIs were how much slower to detect margin deterioration than peers tracking 8–12 metrics?
3. The Miller threshold of 7 ± 2 items, referenced in the lesson's cognitive economics section, originally describes the limit of:
4. Select ALL correct answers about the pyramidal dashboard structure recommended in the lesson.
Sélectionnez toutes les réponses correctes.
5. Select ALL correct answers about why most CFO dashboards fail, based on the lesson's framing.
Sélectionnez toutes les réponses correctes.
Knowing the frameworks is the easy part. The hard part is the political and architectural work of dashboard reduction. Here is the sequence we recommend, drawn from CFO transitions at over 30 mid- and large-cap firms since 2022.
Print every metric currently on your executive dashboard. For each, document:
In our experience, 40–60% of metrics will have no decision attached in the last 12 months. Those are immediate demotion candidates.
Reverse the logic. List the 8–15 actual decisions you make as CFO each quarter: capital allocation, hedging policy adjustments, headcount approvals, M&A go/no-go, dividend recommendations, covenant strategy, etc. For each decision, identify the 1–3 metrics that would change your answer. The union of those metrics is your tier-one dashboard.
This is exactly the process Ruth Porat used at Alphabet when she inherited the role in 2015 and again when she expanded her remit to President & CIO in 2024. Porat's dashboard reportedly fits on a single page and prioritizes capexcapexCapital Expenditure (CapEx) is money spent to acquire, upgrade, or extend long-lived assets like equipment, property, or software that deliver value over multiple years.Voir la définition complète →-to-cloud-revenue ratio, operating margin ex-traffic-acquisition-cost, and a proprietary "AI infrastructure ROIROIReturn on Investment: the ratio of net profit to the cost of an investment. A 300% ROI means each dollar invested returns $3.Voir la définition complète →" metric introduced in 2024.
Structure your reporting in three explicit layers:
The cardinal rule: a CFO should never see a Tier 2 or Tier 3 metric on their primary dashboard. They should *be able* to reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.Voir la définition complète → it in two clicks.
The dashboard will re-bloat. It always does. Build in a quarterly "metric sunset review" where any KPIKPIKey Performance Indicator, a measurable value that shows how effectively you're achieving a specific objective, tracked over time against a target.Voir la définition complète → on Tier 1 must justify its continued inclusion. JPMorgan's CFO Jeremy Barnum has reportedly enforced this discipline since 2022: every new metric proposed for the executive dashboard must come with a paired proposal for which existing metric it replaces. The result: a stable count of 10 metrics for three consecutive years.
A practical complication: regulatory regimes now demand metrics that don't naturally belong on a decision dashboard. Three rules:
1. OECD Pillar Two ETR tracking — keep this on Tier 1, but as a single covenant-style indicator ("ETR vs. 15% floor by jurisdiction, headroom in basis points"), not a 30-row jurisdiction table.
2. CSRD double-materiality KPIs — these belong on a separate sustainability dashboard reviewed quarterly, not the financial control dashboard. Combining them creates cognitive overload without improving either decision.
3. IFRS 16 lease metrics — surface only the *change* in right-of-use asset and lease liability quarter-over-quarter as a single Tier 1 line; the rest goes to Tier 2.
1. Run the 30-day audit this week. Print your current executive dashboard. For every metric, write the last specific decision it drove and the date. Demote anything where you cannot name a decision in the last 12 months.
2. **Conduct a decision inventory before design