SaaSpocalypse revisited: cybersecurity wins, Canva slips, and the survivors nobody expected

SaaSpocalypse Revisited: Which SaaS Survivors Actually Have Moats

Eight months ago, the case for mass SaaS extinction looked airtight. The logic was simple: large monolithic software stacks are less of a competitive moat today than a decade ago, and if every company can vibe code its own CRM, why pay for yours? In February, the hosts declared the thesis canceled—not because the long-term case for AI disruption was wrong, but because the timeline was too aggressive and the analysis too broad.

The numbers bore that out. Roughly $2 trillion in market cap was wiped from SaaS companies during the broader selloff. Much of that has since returned. The iShares Global Tech ETF is up 30% over the past six months alone. But the recovery masks a harsh sorting: some SaaS companies are genuinely unstoppable; others should have stayed buried.

The survivors have sources of strength that code alone cannot replicate.

Shopify is the clearest example. The stock popped 20% after its last earnings report, leading one observer to note that "Shopify isn't a victim of AI—AI is a victim of Shopify." For a business doing nearly $100 million in annual revenue, Shopify costs roughly $1,000 per month—around one-tenth of 1% of revenue. Even with today's AI models, you would need multiple engineers coding around the clock to build a comparable system. Add in the ecosystem of applications already integrated with Shopify, and the math becomes obvious: every hour spent rebuilding Shopify is an hour not spent on revenue-generating work.

The value proposition extends to other network-effect businesses. Spotify and Roblox faced the same doomsday logic—AI can generate music, AI can generate games—yet both have endured. But AI-generated music distributed through Spotify reaches audiences and generates royalties; the same song left on the open internet generates neither. Distribution is its own moat. Roblox owns not just game creation tools but the infrastructure, network, and audience that make those games discoverable.

Marketplace dynamics and go-to-market strength matter more than lines of code.

The hosts identified six case studies six months ago: Google, Meta, Spotify, Shopify, Roblox, and Salesforce. Google and Meta should never have been caught in the SaaSpocalypse panic. Spotify and Shopify proved the thesis wrong. But the real insight lies in what separated winners from victims.

Palo Alto Networks and CrowdStrike are the clearest SaaS winners. Palo Alto is up 121% over the past year and now valued at $320 billion. CrowdStrike is up 107% and worth $230 billion. Cybersecurity has become more critical as AI proliferates, not less. Palo Alto CEO Nikesh Arora bought $10 million of stock five months after the SaaSpocalypse call was canceled. The position is now worth $26 million. (One observer suggested he should have gone 10x levered if he really believed the thesis.)

Some SaaS companies are genuinely obsolete.

Chegg is the canonical casualty. The stock is down 99% over five years. During COVID, Chegg's homework help service was essential. Now, ChatGPT and Gemini answer homework questions for free and walk students through their work. There is no moat. In 2024, Chegg had $617 million in revenue. In 2025, that fell 39% to $376 million. The company is now valued around $80–90 million and will have to cut expenses every year just to eke out profit. Survival requires continuous shrinkage.

Canva is showing signs of sliding into the same trap. The Information reported that Canva's core use cases—designing birthday invitations, infographics, simple graphics—are being displaced by image models like Grok Imagine and others. These models have cracked high-fidelity text generation without misspellings, which was once Canva's defensibility. If you need a one-shot graphic, the model is now faster than logging into Canva.

Infrastructure and tools companies are seeing the opposite effect.

Twilio is up 150% over the past year despite a similar pandemic boom-and-bust trajectory as Chegg. It bottomed around $6 billion and is now a $38 billion company. The difference is structural: Twilio handles SMS routing across mobile carrier networks, spam flagging, and decades of carrier relationships. You cannot vibe code around the complexities of telecom infrastructure. Even if you could theoretically do it, the actual work of maintaining those relationships makes it irrational to try. Twilio becomes infrastructure that agents will simply pull off the shelf.

The same logic applies to data infrastructure. Datadog, analytics platforms, and databases are thriving because every AI company generates exponentially more data and needs exponentially more logging, monitoring, and storage. These companies benefit from the AI wave rather than being disrupted by it.

The sorting is now clear. Businesses with low revenue capture (Shopify at 0.1%), network effects, marketplace dynamics, or infrastructure that is too complex to replace are thriving. Businesses whose core value is answering questions or creating simple designs—tasks that base LLMs now handle natively—are shrinking toward irrelevance. The babies thrown out with the bathwater were real. But so were the actual victims that deserved what they got.