The Cash Machine

The Cash Machine

Salesforce sells customer-relationship software by subscription and, since fiscal 2024, has turned a decade of hyper-growth into a high-margin cash engine: free cash flow reached $14.4 billion in the year to January 2026, up from $1.7 billion nine years earlier, and it has grown every single year in between. The share price tells the opposite story — down 54% from its December 2024 peak. This chapter orients a reader who is new to the name and fixes the question the rest of the report answers.

What the company is

Salesforce is the largest independent maker of customer-relationship management (CRM) software. It delivers its applications over the internet on a subscription basis, and reports as a single operating segment, drawing revenue from two sources: subscription and support, which was roughly 95% of the $41.5 billion total in fiscal 2026, and a small professional-services remainder [1]. Customers pay to access cloud services they never take possession of, and revenue is recognized ratably over the contract term [2].

The subscription base is spread across five product families — the company now brands them all under the "Agentforce" AI label, but the underlying franchises are the familiar ones: Service, Sales, Platform (including Slack), Integration and Analytics (MuleSoft and Tableau), and Marketing and Commerce.

Source: FY2026 Annual Report (Form 10-K), disaggregation of subscription and support revenue [3].

No single family is more than a quarter of the subscription mix, and the two oldest — Service and Sales — still account for the largest shares. Total subscription and support revenue was $39.4 billion, growing 10% year over year [4]. The business grew a large chunk of that footprint by acquisition — Slack, Tableau, MuleSoft, and most recently Informatica — a strategy that shows up on the balance sheet as $58 billion of goodwill, discussed below.

The two curves that define the case

Salesforce's history splits into two lines that diverged in fiscal 2024. Revenue compounded at roughly 19% a year across the decade — from $8.4 billion in fiscal 2017 to $41.5 billion in fiscal 2026 — but for most of that period the company converted almost none of it into reported operating profit, running GAAP operating margins between 2% and 4% while it spent to grow. Then the margin line broke upward: operating margin went from 3.3% in fiscal 2023 to 20.1% in fiscal 2026.

Source: derived from reported financials, FY2017–FY2026 Forms 10-K [5].

The turn was deliberate. Management describes "various restructuring initiatives to improve operating margins," including headcount reductions, as an explicit "commitment to profitable growth" [6]. That pivot — from growth financed at any cost to margin discipline — is the change that reshaped the cash line, and it is the reason that line looks the way it does.

Cash conversion

For a business that recognizes revenue over multi-year contracts and collects much of it upfront, the cash statement matters more than the income statement. Salesforce's is unusually clean. Capital expenditure is trivial — about 1.4% of revenue — because customers run on infrastructure the company largely rents, so free cash flow tracks operating cash flow closely and has risen every year for a decade, through the pandemic, the growth slowdown, and the restructuring.

Source: derived from reported financials, FY2017–FY2026 Forms 10-K, Consolidated Statements of Cash Flows. Capital expenditure averaged under 2% of revenue [7].

Ten consecutive years of rising free cash flow, compounding at about 27% a year, is the kind of consistency a cash-flow-focused investor looks for first. Two cautions belong in the same breath. The gap between the 20% GAAP operating margin and the 35% free-cash-flow margin is largely stock-based compensation, a real cost to shareholders that cash flow flatters — a thread later chapters take up. And the smoothness owes something to the subscription model's prepaid, deferred-revenue mechanics, which the balance-sheet chapter should test rather than take on faith.

What management does with the cash

The company began returning capital in fiscal 2023 and has scaled it quickly. The Board's repurchase authorization now totals $50 billion, expanded by $20 billion in September 2025; in fiscal 2026 alone Salesforce bought back $12.7 billion of stock and, since the program began, has returned close to $27 billion through buybacks and dividends [8] [9]. A dividend, initiated in fiscal 2025, was raised to $0.44 per share declared in February 2026 [10].

Source: FY2026 Annual Report (Form 10-K), Share Repurchase Program and Dividend Policy [11] [12].

The buybacks have started to shrink the share count — from a peak near 997 million in fiscal 2023 to about 923 million by February 2026 [13]. The counter-fact a skeptic should hold onto: the $12.7 billion of fiscal-2026 repurchases were done at an average price of $254.21 [14] — roughly 49% above where the stock trades today. The commitment to shrinking the count is real; the price paid so far has not been opportunistic.

The dislocation

That is the setup a value investor is handed. Over the two and a half years to July 2026 the operating results improved on every axis that matters — revenue up, operating margin rose roughly sixfold, from about 3% to 20%, and free cash flow more than doubled since fiscal 2023 — while the stock did this:

Source: daily price history; fiscal-2026 figure is the 17 July 2026 close of $170.77. December 2024 intraday-close peak was $367.87.

At $170.77 the stock sits 54% below its December 2024 peak. On the roughly 923 million shares outstanding that is a market capitalization near $158 billion; against $14.4 billion of trailing free cash flow, a free-cash-flow yield of about 9.1%, or 8.7% on enterprise value once the modest $7.1 billion of net debt is added — leverage of about half a year's free cash flow.

Source: FY2026 Form 10-K (net debt, shares outstanding [15]); market cap and yields derived at the 17 July 2026 close of $170.77.

Whether that yield is cheap or a value trap turns on durability, and there are two anchors on either side of that question. In its favor: $72.4 billion of remaining performance obligation — revenue already under contract but not yet recognized — sat on the books at year-end, up 14% year over year and equal to about 1.7 years of revenue [16]. Against it: subscription growth has slowed to 10%, the company itself flags that new AI offerings "may disrupt our service offerings" [17], and $58 billion of goodwill means tangible book value is deeply negative — the acquisitions have to keep earning their price.

The question this report answers

A professional investor meeting Salesforce for the first time now has the shape of it: a subscription software franchise with recurring, prepaid economics, an asset-light cash profile, a management team two years into a genuine margin-and-returns discipline, and a stock the market has cut in half. The thread that runs through every chapter to come is this:

Is Salesforce's CRM franchise durable enough — against decelerating subscription growth and an unsettled shift to AI agents — to keep compounding free cash flow, so that the cash it generates a decade from now is reliably larger than the roughly $14 billion it produces today? Everything else — the moat, the deferred-revenue mechanics, the stock-based compensation, the acquisition record, and what the halved price now implies — is evidence for or against that one durability question.