Navigating the SaaSpocalypse

Why AI Kills Weak SaaS but Strengthens Disciplined Vertical Investors

Dougal Cameron

Founder & CEO, Golden Section Growth Equity · B2B Vertical SaaS · February 2026

The global software sector lost $830B+ in six trading sessions. The fear is real. The narrative is incomplete. This is the full picture — market data, attribution analysis, and why disciplined vertical investors are positioned to win.


I Executive Summary

The most severe SaaS selloff since COVID — $830B+ erased in six sessions II Timeline Six months of escalating pressure from silent deceleration to panic selling III Historical Context Every major SaaS drawdown has fully recovered — this isn't unprecedented IV What the Narrative Ignores Three truths the market is missing about vertical vs. horizontal SaaS V Our View Why vertical software with domain expertise is structurally insulated VI AI as Tailwind The same forces creating panic work in favor of disciplined investors VII Attribution Analysis What's actually driving the selloff — AI disruption plus pre-existing conditions VIII The Bottom Line AI kills weak SaaS and strengthens disciplined vertical investors


Executive Summary

–30%
IGV PEAK-TO-TROUGH

$830B+
MARKET CAP ERASED

16.35
RSI — MOST OVERSOLD

The iShares Expanded Tech-Software Sector ETF (IGV) plunged 30% from its September 2025 peak, erasing over $830 billion in market capitalization in just six trading sessions. Jefferies coined the term "SaaSpocalypse" to describe the carnage.

Our central finding: AI disruption fears account for approximately 60–70% of the selloff's severity. However, AI gave the market "permission to finally re-rate what the numbers had been screaming for three years" — persistent growth deceleration, valuation compression, and deteriorating profitability metrics.

The critical distinction: the SaaSpocalypse is overwhelmingly hitting horizontal, general-purpose SaaS. Vertical SaaS with deep domain expertise, regulatory moats, and embedded workflows is structurally insulated — and positioned to benefit.


SECTION II

Six Months of Escalating Pressure

What began as quiet multiple compression accelerated into the most aggressive repricing of the SaaS business model in two decades.

SEP 2025

  • IGV peaks at ~$118
  • All-time high for the software sector ETF. Beneath the surface, growth rates declining since 2021.

OCT–DEC 2025

  • Silent deceleration
  • Quiet multiple compression. Median EV/Revenue falls from 7.3× to 5.1×.

JAN 12, 2026

  • Claude Cowork launched
  • Autonomous AI for non-coding work tasks. Enterprise disruption fears reignite.

JAN 29

  • Worst software trading day since COVID
  • ServiceNow -11% despite beating earnings. Microsoft sheds $360B in a single session.

JAN 30

  • Cowork plugins released
  • Legal, finance, sales, marketing plugins demonstrate autonomous agency across workflows.

FEB 2

  • Palantir Q4 earnings
  • CEO Alex Karp declares AI is "replacing, not augmenting" enterprise software. Revenue +70% YoY.

FEB 5

  • OpenAI unveils Frontier
  • A "Semantic Operating System" treating enterprise apps as data silos for AI agents.

FEB 6

  • IGV closes at $82.46 with +3.5% bounce. Trading volume 37.5M shares — nearly 5× normal.

SECTION III

This Selloff Isn't Unprecedented

Every major SaaS drawdown in the past two decades has fully recovered. The current rout follows a familiar pattern. 2008–09 GFC: Software fell 52% and recovered to new highs within 3 years. 2020 COVID: A 34% decline recovered in 5 months and doubled within 12. 2022 Rate Rout: IGV dropped 30% and recovered to September 2025 highs in under 2 years. "This notion that the software industry is in decline and being replaced by AI is the most illogical thing in the world and time will prove itself." — Jensen Huang, CEO, NVIDIA

Three Truths the Market Is Missing

  • Truth 1: It Is Mostly Horizontal. The damage is concentrated in horizontal, general-purpose SaaS. Vertical SaaS is growing ~32% annually versus ~12% for horizontal — 2–3× faster.
  • Truth 2: Growth Was Already Declining. Public SaaS growth rates declined every quarter since 2021. Median growth fell from 30% to 15%. AI gave the market permission to re-rate what numbers had been signaling for years.
  • Truth 3: Rates Changed Everything. The 10-year Treasury went from 1.5% to 4.5%, re-rating all long-duration assets. Paying 9× revenue at 4.5% reflects stronger conviction than 10× at 1.5%.

Vertical Software Is the Place to Be

  • 85%+ TARGET ENTRY EV/REV
  • TVPI IN PROFITABLE COS
  • ~32% HORIZONTAL GROWTH

Domain expertise, regulatory moats, workflow depth, and value-based pricing create structural insulation that generic AI cannot breach.

At-Risk SaaS:

  • Pure automation targets with simple admin workflows. Low switching costs and thin workflow wrappers. Seat-based pricing without usage migration. Mid-market horizontal SaaS. Document and content tools easily replicated by AI.

Strengthened SaaS:

  • Core workflow orchestration in regulated verticals. Deep data moats and proprietary domain data. High switching costs and compliance requirements. AI-enablement layers embedded in vertical workflows. Usage-based and outcome-based pricing models.

AI as Tailwind

As capital retreats from software broadly, competition for quality vertical deals declines. AI lowers barriers to build v1, but v1 is 2% of the work. Portfolio companies embedding agentic features become AI-native cooperators, not casualties.

AI-driven efficiency gains compress R&D, improve margins, and improve Rule of 40. 85%+ of TVPI is in profitable companies.


SECTION VI–VII

What's Actually Driving the Selloff

AI disruption is real — but it ignited a fire in a forest that had been drying out for three years.

  • Direct AI Disruption: 60–70%
  • AI-Accelerated Trends: 20–30%
  • Non-AI Factors: 10–20%

The fear: "AI kills SaaS."
The reality: AI kills weak SaaS.
And strengthens disciplined vertical investors.

Vertical software is structurally insulated.


Interested in our research?

Golden Section publishes research on market dynamics, vertical SaaS, and the intersection of AI and enterprise software.