Long-term value metrics: how the best companies think beyond quarterly EPS

In February 2024, Unilever's then-incoming CEO Hein Schumacher quietly killed off the "Sustainable Living Plan" branding that Paul Polman had spent a decade building, but kept almost every underlying long-term metric. The market briefly cheered what it read as a return to "discipline." Eighteen months later, activist Nelson Peltz was still on the board, the share price had underperformed Procter & Gamble by roughly 22 percentage points over five years, and Schumacher's own strategic review concluded that Unilever's *real* problem was that it had been managed to too many short-cycle financial targets, not too few long ones. The lesson buried in that episode is the one this module is built around: companies that confuse cadence with control end up optimizing the wrong variables, and the CFO is the person who decides which variables sit on the dashboard.

This lesson examines how the most disciplined long-term operators, Amazon, Berkshire Hathaway, Constellation Software, and a handful of European industrials, **construct KPIKPIKey Performance Indicator, a measurable value that shows how effectively you're achieving a specific objective, tracked over time against a target.View full definition → architectures that *predict* future cash generation rather than *report* past accounting outcomes**. We'll move from philosophy to the specific metrics you can install in your management reporting pack by next quarter-end.

Why quarterly EPS is a lagging, manipulable, and increasingly dangerous anchor

Start with the empirical case. Research from McKinsey's Corporate Performance Analytics group (updated through 2025) consistently shows that roughly 70-80% of a typical large-cap equity's value is attributable to cash flows expected beyond three years. Yet the median S&P 500 earnings call in 2025 spent over 60% of management commentary on the *current and next* quarter. The mismatch between where value lives and where management attention sits is not a rounding error, it is the structural pathology of public-company finance.

EPS itself is a particularly poor long-term compass for three reasons that any CFO operating under IFRSIFRSThe global accounting rulebook that governs how companies report financial results, used across the EU and 140+ jurisdictions.View full definition → 16, OECD Pillar TwoOECD Pillar TwoOECD-backed rules imposing a 15% minimum effective tax rate on large multinational groups, jurisdiction by jurisdiction.View full definition →, and CSRDCSRDEU directive requiring large companies to report standardized, audited sustainability data alongside financial results.View full definition → should now feel acutely:

• It's increasingly distorted by accounting geography. IFRS 16 moved operating leases onto the balance sheet, restructuring 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.View full definition → upward and depreciation/interest downward in ways that vary wildly by industry. Pillar Two's 15% global minimum tax, fully effective across the EU, UK, and most OECD jurisdictions in 2024-2025, has compressed effective tax rate arbitrage and created top-up tax volatility that scrambles year-on-year EPS comparability for multinationals.
• It's manipulable through buybacks at a scale that now dwarfs organic earnings growth. S&P 500 buybacks ran at roughly $900 billion in 2024. A flat-net-income company can deliver 4-6% EPS growth purely through share count reduction, and the market, embarrassingly, often rewards it.
• It ignores the cost of the capital that produced it. This is the original Stern Stewart critique from the 1990s, and it remains correct: net income that doesn't clear the weighted average cost of capitalweighted average cost of capitalThe blended rate a company pays to finance itself through debt and equity. It sets the minimum return an investment must clear to create value.View full definition → destroys value even as it "beats consensus."

The bezos doctrine, in practice

Jeff Bezos's 1997 shareholder letter, which Andy Jassy has reprinted with every annual report since taking over as CEO in 2021, contains the cleanest articulation of a long-term metric philosophy in corporate America: *"When forced to choose between optimizing the appearance of our GAAPGAAPThe standard set of accounting rules companies follow to prepare consistent, comparable financial statements, dominant in US reporting.View full definition → accounting and maximizing the present value of future cash flows, we'll take the cash flows."*

This is not rhetoric. Amazon's internal management reporting, as described by former CFO Brian Olsavsky in multiple investor settings, is built around free cash flow less equity-based compensation and lease principal repayments, a metric Amazon began disclosing prominently in 2020 because it captures the economic reality that SBC is a real cost and leased fulfillment centers are real capital. In FY2024, Amazon reported approximately $38 billion of this adjusted FCFFCFFree Cash Flow is the cash a company generates from operations after funding the capital expenditures needed to maintain and grow its asset base.View full definition → figure against roughly $59 billion of GAAP operating cash flow, a deliberate $21 billion downward adjustment that most CFOs would have buried.

The takeaway: **Amazon's KPI architecture is engineered to make management *uncomfortable*, not comfortable**. That is the design principle most finance functions get backwards.

Building a long-term value KPI architecture: the four-layer model

A defensible long-term KPI stack has four layers. Most finance teams stop at layer two.

Layer 1: economic profit, not accounting profit

The foundational metric for any long-term system is some form of economic profit, NOPAT minus a capital charge. Call it EVA, residual income, or "cash value added." The name matters less than the discipline.

The exemplar here is Constellation Software, the Toronto-listed vertical-market software acquirer run by Mark Leonard. Constellation publishes one of the most candid annual letters in the industry. Their entire capital allocation engine is governed by a single internal hurdle: every acquisition must clear a defined IRR threshold (historically disclosed as in the 20%+ range, though Leonard has publicly noted in his 2022 and 2023 letters that hurdle rates had to be lowered as scale increased). The compensation of operating-group presidents is tied to a metric Constellation calls "ROIC + organic net maintenance revenue growth." It is essentially economic profit, expressed in growth-adjusted form.

Result: from its 2006 IPO through end-2024, Constellation compounded shareholder returns at roughly 33% annually. The KPI did the work.

Layer 2: cash conversion and reinvestment quality

The second layer asks: of the economic profit you generate, how much converts to deployable cash, and what return are you earning on incremental investment?

Two metrics worth installing:

• Cash conversion ratio: FCF / Net Income, tracked on a five-year rolling basis. Consistently below 80% is a yellow flag for earnings quality.
• ROIIC (Return on Incremental Invested Capital): the change in NOPAT divided by the change in invested capital over a 3-5 year window. This is the single best predictor of whether reinvested earnings are creating or destroying value.

Warren Buffett's 2024 Berkshire letter once again highlighted that he evaluates operating subsidiaries on per-share intrinsic value growth over rolling five-year periods, explicitly noting that any single year, including 2023, when Berkshire posted record GAAP earnings inflated by mark-to-market gains on its Apple stake, is "essentially meaningless." Charlie Munger before his death in November 2023, and Greg Abel after, have publicly framed the Berkshire metric as: *would a rational private owner pay more for this business today than five years ago, after accounting for cash extracted?*

Layer 3: leading operational indicators

This is where most CFOs underinvest. The financial metrics in layers one and two are still lagging, they describe outcomes. **Leading indicators describe the *drivers* of those outcomes 6-24 months out**.

The specific leading indicators differ by business model, but the canonical examples:

• Subscription/SaaS: Net Revenue Retention, CAC payback in months, magic number, Rule of 40
• Industrial: Book-to-bill ratio, backlog quality (margin-weighted), aftermarket attach rate
• Consumer: Household penetration, repeat purchase rate, gross margin per active customer
• Capital-intensive: Capacity utilization at peak vs. average, maintenance capex as % of replacement cost

Layer 4: intangible and stakeholder capital

Under CSRD, which began phased reporting for large EU and EU-active companies in FY2024 disclosures, intangible and stakeholder metrics are no longer optional commentary, they are audited disclosures. The smart CFOs are using this regulatory push to install metrics they actually want to manage by, rather than treating CSRD as a compliance burden.

The metrics that have predictive power:

• R&D productivity: revenue from products launched in the last 3 (or 5) years as a percentage of total revenue. 3M historically targeted 30%+; the decline of this metric from 33% in 2010 to under 25% by 2020 prefigured 3M's lost decade.
• Employee engagement and regretted attrition: not the HR vanity score, but specifically the attrition rate of top-quartile performers. Microsoft under Satya Nadella has reportedly tracked this monthly since 2015.
• Customer lifetime value to CAC ratio, refreshed annually with actual cohort data, not modeled assumptions.

From architecture to operating cadence: what changes on monday morning

Designing the metric stack is the easy part. The harder part, and the part that separates CFOs who actually change behavior from those who merely publish new dashboards, is rewiring the operating cadence around long-term metrics.

Case study: how DSM-firmenich restructured its management reporting

The 2023 merger that created DSM-Firmenich provided an unusual natural experiment. CFO Ian Kelly, appointed in 2024, inherited two different reporting traditions: DSM's classic European EBITDA-centric model and Firmenich's privately-held, cash-and-IRR-centric model. Rather than averaging them, Kelly publicly committed to a reporting architecture organized around three time horizons:

• Quarterly: organic growth, adjusted EBITDA margin, cash conversion, the minimum disclosed to the market
• Annual: ROIC, free cash flow per share, innovation revenue (% from products launched in the last 5 years)
• Three-year rolling: economic profit, intrinsic value per share (calculated and disclosed internally to the board)

The critical move: executive compensation was rebalanced so that 60% of long-term incentive value vests against the three-year metrics, not relative TSR. This is the structural change that makes the architecture credible. As Kelly put it in the company's 2024 Capital Markets Day, "If the comp committee doesn't believe the metrics, neither will the operating teams."

The three practical shifts a CFO must make

For the CFO building this on Monday morning, three concrete shifts:

First, separate the disclosure pack from the management pack. Your quarterly investor disclosure should comply with consensus expectations and SEC/ESMA requirements. Your internal management pack, the one the executive team actually runs the company by, should lead with ROIC, ROIIC, FCF per share, and 2-3 industry-specific leading indicators. The mistake is letting the investor pack become the management pack.

Second, change the comp plan, or admit you're not serious. If LTI vests on one-year relative TSR (as it does at an embarrassingly large share of S&P 500 companies), no amount of long-term rhetoric will change behavior. The defensible structure in 2026, given Pillar Two's compression of tax arbitrage and CSRD's elevation of non-financial metrics, is LTI vesting over 3-5 years against a basket including ROIC, cumulative FCF, and 1-2 strategic non-financial measures.

Third, install a five-year rolling review at the board level. Berkshire does this. Constellation does this. Most boards spend