For the better part of two decades, SaaS was the greatest business model Wall Street had ever seen. Low marginal costs, predictable recurring revenue, and a customer base sticky by design — not because the product was irreplaceable, but because switching hurt. Investors rewarded this with multiples that defied gravity. 20x revenue was normal. 40x wasn't shocking. The model printed money, and the market priced it in forever.
That era is quietly ending. And the vendors know it.
The Math That Made SaaS Magic
The SaaS multiple was always a bet on the cost structure. Once the software was built, adding a new customer was nearly free. Gross margins of 70–80% were the norm. The pitch was simple: suffer losses now, lock in customers, then watch the incremental economics become extraordinary.
It worked — until the thing that made software expensive to build stopped being expensive.
AI has collapsed the capital required to create software. A small team with the right tools can now build in weeks what once took years and millions of dollars. The moat that justified those multiples — the sheer cost and complexity of replicating a mature SaaS product — is eroding in real time. Every quarter, the cost to build a credible alternative drops. Every quarter, the multiple the incumbent deserves drops with it.
The Panic Is Visible — And It's Coming Out of Your Pocket
SaaS companies are not sitting quietly with this reality. They are acting on it, and their customers are footing the bill.
Prices are going up. Multi-year contract extensions are being pushed aggressively — not because they offer customers better economics, but because they buy vendors time. Locking in three or five years of revenue smooths the story for investors while the underlying model is quietly called into question. The urgency your account rep suddenly has around "locking in current pricing" is not a favor. It's a survival tactic.
At the same time, vendors are racing to bolt AI features onto existing products — not because customers demanded them, but because the narrative demands it. An "AI-powered" rebrand buys time on earnings calls. It justifies price increases on contracts already under scrutiny. But most of these additions are cosmetic. A summarization button here. A chatbot sidebar there. Features shipped to protect a multiple, not to solve your problem.
The Data Trap Is the Real Play
Beneath the AI theater, something more deliberate is happening. SaaS vendors understand that the product itself may no longer be defensible. So the new moat is your data.
The strategy is to make your data as entangled in their ecosystem as possible — difficult to export, impossible to fully own, and conveniently most useful inside their walls. Each platform wants to be your AI headquarters, training its models on your slice of data, building agents and copilots that operate only within their environment. The result isn't intelligence built for your business. It's intelligence built to retain your subscription.
And here's the absurdity of it at scale: your company is now paying for six, eight, twelve different SaaS tools, each with its own AI layer, each hoarding its piece of your data, none of them talking to each other. Not because that architecture serves you — but because cooperation between vendors doesn't serve them.
Whose Roadmap Are You Actually On?
This is the question every technology leader needs to sit with. When your vendor makes a product decision, launches a new module, deprecates an old one, or prices an AI feature into your next renewal — whose interests drove that decision?
The answer is almost never yours. It is the vendor's margin, their investor narrative, their competitive positioning. You are a revenue line in a model that is under pressure. The decisions being made on your behalf are being made by companies scrapping for a financial future, not designing for your operational one.
There is no version of this dynamic that reliably leads to your best outcome. Design by vendor is not a technology strategy. It is a delegation of your most critical decisions to parties with misaligned incentives.
The Kicker: Own Your IP
Companies that will navigate this transition well share one trait — they have invested in owning their own technology, data, and roadmap.
This doesn't mean building everything from scratch. It means being deliberate about where you accept dependency and where you don't. It means treating your data as a strategic asset that lives with you, not with your vendor. It means building or acquiring capabilities that reflect your business logic — not licensing a generic approximation of it.
The era of cheap, abundant software creation is an opportunity, not just a threat. You can now build what you actually need at costs that were unimaginable five years ago. But only if you're not already locked into a multi-year contract that was signed by someone who didn't see this coming.
Understand the motivations of every party at the table. Some vendors are no longer your partners as they once positioned themselves to be. Know what they need — and then make sure what you build serves what you need, independent of whatever roadmap they run down next in their pursuit of margin.
The SaaS multiple is a ghost. Don't let it haunt your balance sheet, too.
.png)
.png)
.png)
