Álvaro de Nicolás
← All essays

Musings · SaaS & capital markets

SaaS, AI and the Next Cycle

Personal musings · Álvaro de Nicolás

Some unpolished thoughts on where SaaS is actually going. Less a thesis, more a set of observations I keep returning to in board conversations.

1. Big-cap SaaS is being repriced in public

One of the largest names in the category recently lost roughly $50B in market cap on a single growth deceleration. That is not noise. It is the market saying: the multiples we paid for steady-state SaaS growth required a future that no longer looks guaranteed. When that big a tanker turns, every smaller boat moves.

2. The PE write-down nobody wants to talk about

A $3.5B private-equity acquisition in the developer-skills space was recently written down to zero. Not down 30%. Not impaired. Zero. That is one of the most violent signals you can get about a category, a thesis, and a vintage. It tells me three things:

"Recurring" was a financial label. "Durable" is an operating verdict. AI is collapsing the gap between them.

3. The IPO window is frozen — and that has consequences

The exit market has been closed long enough that it is shaping behaviour. Founders are tired. Boards are tired. Secondaries are doing more work than they were designed for. When the window finally opens, the first cohort through will not be the loudest — it will be the ones with clean unit economics, defensible data, and AI inside the product, not bolted on.

4. Hyper-growth privates vs. struggling publics

The most interesting split right now: a handful of private hyper-growth names continue to compound at rates the public comp set cannot touch, while several once-darling public names are being punished for missing by a single point. The market is rewarding category leadership and punishing category participation. There is no middle.

5. AI is not a feature line — it is rewriting the budget

The most underappreciated CFO conversation of this cycle: AI is not showing up as a new line in the IT budget. It is showing up as a re-allocation of the existing one. Seats are being deferred. Renewals are being rescoped. Token spend is moving from R&D experiments into production cost of goods. Companies that still pitch "AI on top of your current stack" are going to be priced as add-ons. Companies that re-architect the workflow will be priced as platforms.

6. The arms race is real, but the moat has moved

The competition between the frontier labs is genuine and consequential. But for operators, the moat is no longer "which model do you use." It is:

If the answer to any of these is "we'll figure it out," you are building on rented land.

What I am telling boards

None of this is a prediction. It is a posture. The cycle will do what cycles do. The question is whether your operating model is built to compound through it, or to defend against it.