Preface: Why Software Is Still Worth Studying
In the winter of 2021, open any technology strategy note from any Wall Street bank and you would find something close to an article of faith. Cloud software — what we now simply call SaaS — was described as a business that could barely make a mistake. Customers paid by the year and paid more the longer they stayed; serving one more of them cost the company almost nothing. Revenue compounded on itself like a snowball, and valuations only went up. That year, the market was willing to pay twenty, thirty, sometimes more times a high-growth software company’s annual revenue in enterprise value. People had stopped asking whether a company could make money and started asking only how fast its revenue could still grow. Software, it seemed, had finally become a compounding machine that never had to stop.
Then interest rates arrived.
In 2022 the Federal Reserve raised rates at a pace not seen in forty years. For a compounding machine, rates are the most quiet and most dangerous adversary: when the risk-free yield climbs from near zero back above 4%, the mathematical foundation under any company that has pushed its profits far into the future is pulled away in an instant. The market’s whole logic for pricing software was rewritten in a matter of months. Multiples of twenty and thirty times revenue collapsed into the single digits; a roster of marquee SaaS names fell by half from their highs, then by half again. More important, investors began — for the first time in years — to press on a question the growth story had long papered over: strip out the stock compensation, the acquisitions, and the one-time billing timing, and how much real cash has this company actually created for its shareholders?
This book begins with that question.
One valuation reckoning, three neglected truths
On the surface, the 2021–2022 reckoning was a rate-driven repricing. But it tore open three layers of truth that had rarely been taken seriously — and those three layers happen to form the skeleton of this book.
First, “recurring” revenue is not automatically “high-quality” revenue. The subscription model makes revenue predictable, but predictable revenue is not necessarily revenue that can last, expand, and turn into profit. A company that loses a fifth of its customers every year and can only sustain growth by burning cash to acquire new ones has “recurring revenue” that is nothing like the recurring revenue of a company whose customers stick and expand on their own — even though, on the top line of the income statement, the two look identical.
Second, a beautiful gross margin is not the same as real shareholder return. Software companies often post gross margins of 70% or 80%, which makes them look like money printers. But beneath that margin can sit staggering sales costs, operating losses that never turn positive, and enormous compensation paid in stock rather than cash. That last item matters most: stock-based compensation never shows up as a cash outflow, yet it genuinely dilutes every existing shareholder’s stake. A company whose free cash flow looks robust deserves a large question mark over the quality of that cash flow if the cash was “saved” by continuously issuing new shares.
Third, a software business’s moat depends far more on the position it occupies than on the features it ships. The finest point solution can be erased in a stroke when a platform giant folds the same capability “for free” into a suite; the most developer-beloved product may never convert into a contract an enterprise will keep paying for. Who controls the entry point to the workflow, who holds the data, who owns the channel of distribution — these usually decide the long-run outcome more than who has the better product.
In the most euphoric years of the cycle, these three truths were drowned out by the noise of the growth narrative; in the cold snap of the rate reversal, they were mercilessly lit up again. Understanding them is where any understanding of the SaaS business has to start.
So why is software still worth studying?
By now you might ask: if SaaS has been through a bubble, a slowdown, and a valuation reckoning, why write an entire book to study it?
Because after all of that, software did not exit the stage of history — it embedded itself more deeply than ever in how the modern economy runs. Today almost no company of any scale operates on anything other than layer upon layer of software: customer relationships run on a CRM, people and payroll on an HCM, the books on financial software, data on a data cloud, security across an entire platform of it. Software is no longer a tool that companies buy to become more efficient; it is the operating system on which they run. A business that determines how the world’s companies function, and is still growing at double digits, is worth studying seriously no matter how its valuation rises and falls.
More important, software is standing at the threshold of the deepest transformation since its birth: artificial intelligence.
Over the past two years — from the generative-AI wave that ChatGPT set off to the AI agents debated endlessly today — AI has been putting two opposite questions to SaaS at the same time. On one side, AI may be software’s next growth curve: it lets software do more, command higher prices, open new markets. On the other, AI may be a structural threat to the very foundation of the SaaS business model. If an AI agent can replace a great deal of work once done by people, does pricing “per seat” still stand? If every AI call consumes real money in compute, does the high-margin myth of “near-zero marginal cost” still hold? If the agent becomes the new entry point between people and software, will the applications that own the interface today someday be reduced to pipes — called in the background by an agent, stripped of their pricing power?
These questions have no ready answers. They are exactly what Part Four sets out to explore — and they are the intellectual high ground of the book, because they concern not the past of SaaS but its future.
What this book offers you is a method, not a list
Let me be clear at the outset: this is not a stock-picking book.
It will not tell you which software stock to buy right now. The reason is simple: any specific list of recommendations goes stale the moment prices, earnings, and the competitive map move — and, more fundamentally, talking about a “good company” divorced from its valuation and the expectations already priced in is one of the most common traps in all of investing. An exceptional company whose price already discounts a decade of optimism can be a terrible investment; a mediocre or even declining company can hide an opportunity when the market’s pessimism has overshot. A good product, a good company, and a good stock are three different things — a line that recurs throughout this book, because it is the watershed between understanding a company and understanding an investment.
What the book genuinely means to hand you is a research framework you can use on your own, carry from one company to the next, and keep updating: how to break down where a SaaS company’s revenue actually comes from; how to read, in the seams of the financials, the operating truths management did not spell out; how to weigh growth against retention, profit, and cash flow; how to value a company and think through exactly what expectations its current price implies; and how to judge whether AI, for one specific company, thickens the moat or thins it.
Learn that framework and you no longer need anyone’s list. You can research any software company yourself and reach your own conclusion — one you can test against concrete operating metrics. That is what this book hopes you walk away with.
On the book’s data and its boundaries
So that every figure in the book is internally consistent and can be checked, the book uses a single data cutoff: June 30, 2026. All share prices, market capitalizations, and valuation multiples are as of that date or an explicitly stated retrieval date; each company’s operating data uses the most recent fiscal quarter disclosed as of then (fiscal years end in different months for different companies, and the book flags the basis case by case). Valuations and fundamentals in software move every day; the numbers here are a snapshot of that moment, and should be read as such.
On sources, the book gives priority to primary material: the 10-K annual reports, 10-Q quarterly reports, and 8-K current reports companies file with the U.S. Securities and Exchange Commission (SEC); the earnings calls and shareholder letters on their investor-relations sites; and official product and pricing pages. Only where these primary sources cannot reach does the book cite authoritative market-research firms and high-quality financial media, with cross-verification. Any metric the book calculates itself is shown with its formula and assumptions; anything that cannot be verified is marked “uncertain” or dropped outright; any conclusion that is the author’s judgment rather than fact is set apart with phrasing like “in the author’s judgment” or “this suggests.” This book would rather admit what it does not know than dress a guess up as a fact.
On scope, the book mainly covers SaaS and cloud-software companies listed in the United States, including a few that are headquartered elsewhere but trade on U.S. exchanges. For companies that earn revenue from a mix of subscription, usage-based billing, payment take-rates, advertising, or hardware, the book takes their business models apart rather than treating a company as SaaS in its entirety just because it calls itself a “cloud platform.”
Finally, a necessary reminder: this book offers research methods and analytical frameworks, not investment advice directed at any individual. Markets carry risk, and every investment decision should be made by you, on the basis of your own thorough research.
The road this book will take you down
The book is in five parts. Part One returns to the origin of the subscription economy — why software moved from the disc to the cloud, and how the compounding machine was built. Part Two goes into the financials, teaching you to see a SaaS company’s true quality: the quality of its revenue, the substance of its cash flow, and the danger signals in the financials that are easy to miss. Part Three unfolds a map of the industry, walking you across the battlegrounds of horizontal platforms, data, security, vertical industries, developer tools, and small-business software. Part Four — the center of gravity — digs into how AI is redefining the software business: whether it is a sustaining innovation or a disruptive threat, and how it will rewrite margins, pricing, entry points, and competition. Part Five gathers every earlier framework into an operational research system: how to value, how to tell a good company from a good stock, and how to build and keep iterating a research process of your own.
Now, let us begin with a disc about to go obsolete, and watch how the software business was rewritten from the ground up.