Aki's essay hit twice. June 2024, then again in early 2026 when AI hype peaked. Same piece. Same core argument. Same viral moment. The numbers say the second wave hit harder... and that matters more than the argument itself.
This is not about whether SaaS is actually dying. It's about why we need it to die every 18 months.
1. The Pattern Recognizes Itself
Every disruption narrative follows the same arc. Technology X threatens Business Model Y. Founder tweets apocalypse. 47 substack takes materialize. Major publication runs the contrarian counter-take. Market shrugs. Quietly, the transition happens anyway... usually slower and messier than anyone predicted.
SaaS isn't dying. Certain SaaS business models are compressing. The difference matters. $800/month project management tools are bleeding to free tiers and self-hosted alternatives. But Stripe? Figma? Cloudflare? Still collecting checks. The story sells better when you delete the second half.
2. AI Changed What "Death" Means
In June 2024, the apocalypse narrative was about margin compression. Open source eating lunch money. Consolidation. By 2026, it became about obsoletion... the idea that AI agents render entire categories pointless. That's a different fear. Scarier one. Which is exactly why it resurfaced.
The numbers don't support it yet. SaaS revenue stayed flat-to-up. But the emotional valence shifted. We went from "SaaS is struggling with profitability" to "AI will erase the need for SaaS." One is a market problem. One is existential. The second one spreads faster.
3. Viral Takes Need Freshness, Not Truth
The piece didn't blow up the second time because market conditions changed. It blew up because a new cohort of readers hadn't seen it. The algorithm doesn't know the essay is two years old. It sees engagement spike and assumes novelty. Readers share it thinking they found something hot. The creator benefits from the appearance of prescience.
This is how narrative cycles become self-perpetuating. Each resurgence feels like validation of the original thesis, when it's really just proof that new people exist who weren't paying attention the first time.
4. We're Bad at Distinguishing Compression From Collapse
SaaS margins compressed. That's real. Customer acquisition costs stayed high. Retention got harder. But "compression" doesn't read. "Death" does. So founders who saw revenue flatten or growth slow adopted the apocalypse framing... not because the market was ending, but because it was the only language that felt proportional to their pain.
The rhetoric is accurate about the experience. Just not about the market. Your specific SaaS product might be in trouble. The category isn't. This distinction gets lost in the narrative. It has to. Articles that say "some SaaS is struggling while other SaaS thrives" don't go viral. They're useful. They're honest. They bomb.
5. The Cycle Accelerates Because Stakes Are Unclear
Traders need clarity on what's actually changing. Investors need to know where capital goes next. Founders need to decide whether to pivot or double down. None of them get answers from apocalypse narratives... but they still read them. Because the appearance of explanation beats the silence of genuine uncertainty.
The second viral moment of the "death of SaaS" essay happened because AI hype created fresh uncertainty. The narrative provided a framework. It didn't matter that the framework was recycled. It mattered that it was there.
What Actually Changed
Between June 2024 and 2026... almost nothing structural. Some SaaS companies got leaner. Some failed. Others raised at higher valuations. The underlying dynamics stayed the same: margin pressure, consolidation, market bifurcation between winners and the rest.
What changed was who was paying attention and what they were afraid of. That's the real story. Not whether SaaS dies. Whether we've learned to tell the difference between market shift and market collapse.
The numbers say we haven't.