Surviving the SaaS repricing

LAST REVIEWED 2026-08 · SOURCED FROM 1 SESSION, APR 2026

TL;DR

Public markets have repriced software: roughly $2T of SaaS market cap evaporated in early 2026, and one major vendor’s revenue multiple compressed from ~18x to ~5x even as revenue grew. Three forces drive it: buyers can vibe-code internal replacements, systems of action are eating systems of record, and seat-based revenue collapses wherever the “user” becomes an agent. The strategic response: accept that product is now the easy part, and rebuild your GTM for buyers who ask an AI before they ever visit your site.

What the market is saying

The session opened with the tape: about $2 trillion of public SaaS market capitalization gone in a two-month window, and the flagship example — a major marketing-software vendor that grew revenue from $1.3B to $3.1B while its multiple compressed from 18.6x to 5x. The market isn’t pricing current revenue; it’s pricing the belief that customers can now build the replacement in-house.

“Traditional SaaS is dead… product has become very, very attainable. What’s still extremely difficult is GTM.”

— Multi-exit founder, companies scaled past $75M ARR · session, Apr 2026

The three threats

1. Vibe coding. Buyers assemble internal clones of scheduling tools, project trackers, and dashboard products in days. Anything whose value is a nice UI over a database is exposed.
2. Systems of action beat systems of record. Nobody wants to pay five figures a year for a database UI plus a separately priced automation add-on. Products that do the work are replacing products that store the record — and they don’t price per seat.
3. Seat collapse. Where agents do the work, seat counts fall even as usage grows. The large vendors’ seat revenue is flattening while their usage-priced agentic lines grow.

One honest caveat aired in the Q&A: an enterprise attendee pushed back that idea-to-production in days doesn’t hold in regulated environments — and the speaker’s own government deployment required a federal compliance authorization with a two-year wait and embedded humans. The flip side: that pain justified deals an order of magnitude larger than the commercial equivalent. Regulated markets are slower and more defensible for the same reason.

The playbook that falls out

• AEO over SEO. Buyers now research in AI assistants before ever visiting your site. Optimize to be recommended — presence in the sources assistants cite, structured product information, opinionated comparison content.
• Ship API-first and expose your product to AI tools. The session’s example: an upstart email API eating an incumbent because developers ask their coding assistant, not the docs — and the upstart is what the assistant reaches for. “Instead of the customers logging in to your product, the product actually comes to you” — inside the assistant.
• Agentic onboarding. Activation that took a customer-success team months can happen in minutes when an agent configures the product from the customer’s own data. Time-to-value is becoming the whole onboarding budget.
• Agentic retention. Churn agents that detect usage drops and reach out off-product, expansion agents that propose upsells at the moment of need — retention becomes a product surface, not a staffed function.
• Rebalance the funnel budget. Lead generation that cost ~$100 per lead through agencies runs near $1 with an agent stack; the human budget moves to the parts that still need humans — closing five-figure-plus deals.

“One salesperson with an AI skill is actually a lot more valuable than maybe a 40 BDR team.”

— Same session, Apr 2026

What this means at seed stage

Connect the threads from the investor sessions: if product is cheap, defensibility must come from somewhere else — which is why the fund GPs keep pointing at data moats, and why the traction bar is shifting from “built it” to “distributed it.” Your pitch’s hardest question is no longer “can you build this?” but “why do you win distribution?”

Sources

One session (Apr 2026) with a multi-exit founder — prior company acquired by a major security vendor, products scaled past $75M ARR — now building GTM tooling. Recording is captions-only; quotes are from the featured speaker. Market figures are as presented in the session; verify current numbers before citing.