The software selloff has everyone talking about apocalypse. But I think we’re actually looking at a textbook wave of disruption—and if you understand the waves, you can separate real structural decline from temporary panic.
Here’s what I mean: the panic about AI killing SaaS is partly real, but it’s unfolding in predictable waves. Incumbents don’t die in wave one. They usually get displaced in wave three, after they’ve had a chance to adapt, prove they can innovate, and then fail to outrun purpose-built natives.
The Three Waves
Wave One (now, 6-18 months): “Can AI agents do this?” Fear. Indiscriminate selling. Investors run from anything that looks replicable by Claude or an LLM. The companies getting hit hardest are those with weak moats—commodity workflow tools, expense tracking, basic CRM. But even strong companies get crushed in the panic. Valuation multiples compress because duration is being repriced and structural threat narratives are amplified.
Wave Two (1-3 years): “Which vendors actually solved this?” Reality sets in. Enterprises figure out their build-versus-buy calculus. Some incumbents adapt smartly—they embed AI into workflows, deepen defensibility, use their data advantage. Others bolt features onto legacy architectures and stay vulnerable. AI-native startups gain traction but run into go-to-market friction. Winners and losers become visible. Stock prices stop being driven by existential fear and start reflecting actual execution.
Wave Three (3-5 years): “Where’s the best unit economics?” This is where incumbents usually lose. AI-native companies built from the ground up for this paradigm have lower customer acquisition costs, better margins, faster iteration cycles. Legacy SaaS faces margin compression as it’s forced to compete on price or gets displaced on UX and speed. Most incumbents either get acquired, become “legacy infrastructure” (profitable but slow-growing), or genuinely reinvent early enough to stay competitive.
Wave Four (5+ years): Consolidation. The survivors are either genuine natives who won wave three, or rare incumbents that pulled off a true reinvention.
History Rhymes
I don’t think this is speculation. The pattern repeats.
Electricity displacing steam (1890s-1920s): Steam-powered factories didn’t disappear overnight. But electricity enabled fundamentally different ways to organize work—assembly lines, distributed motors, 24/7 production. Companies that owned steam infrastructure sometimes pivoted. Most became margin-squeezed legacy players. The real winners weren’t steam companies switching to electric; they were companies that understood electricity enabled new business models entirely. Ford didn’t displace Watt; he understood the assembly line.
Newspapers to internet (1995-2015): Newspapers tried to adapt. They built websites, paywalls, digital editions. Some survived (WSJ, FT, NYT—premium brand plus paywalls worked). Most died. The real winners weren’t legacy newspapers going digital. They were internet-natives: Google understood search, Amazon understood e-commerce, Facebook understood social networks. Newspapers tried to out-innovate the internet and mostly failed because the internet wasn’t competing on features; it was competing on business model.
On-premise software to SaaS (2000-2015): This is probably your best parallel. Siebel, Peoplesoft, Bea Systems, Hyperion—all dominant in their domains. Salesforce came in and said: “We’ll host CRM in the cloud, lower your TCO, faster adoption, better UX.” Legacy vendors tried to build SaaS versions (Oracle pushed Fusion, SAP struggled with Success Factors). Some survived, but real winners weren’t on-prem companies that pivoted. They were SaaS-natives that rethought product architecture for multi-tenant, cloud-first design from day one. You can’t win by strapping a SaaS wrapper onto your on-prem codebase.
Intuit: A Case Study in Wave Dynamics
Intuit is the perfect example of how this plays out, because it has real defenses and real vulnerabilities.
Wave One (now): Intuit stock got crushed. Down 45% YTD, caught in the broad SaaS panic. The fear is simple: “AI can generate tax forms, calculate deductions, spot audit risks. Why do you need Intuit?” But Intuit actually has genuine moats: tax code is brutally regulated (IRS compliance, 50 state rules, audit trails), they have 100M+ customers, and 40+ years of tax data. That’s real friction. So Intuit survives wave one. It’s oversold.
Wave Two (1-3 years): Intuit probably improves here. They use AI to deepen defensibility—better refund predictions, faster filing, compliance guarantees backed by data. They have the resources to innovate. Their stock probably recovers as investors realize “okay, Intuit isn’t getting displaced tomorrow.” They might even grow faster as they become more efficient at tax prep and acquire market share.
Wave Three (3-5 years): This is where it gets interesting. An AI-native tax software company—probably founded by someone who got frustrated with Intuit, maybe an ex-Intuit engineer or Stanford AI researcher—ships something purpose-built for this era. It has:
- Lower customer acquisition cost (viral, because it’s 10x easier to use)
- Better unit economics (built for AI orchestration from day one, not retrofitted)
- Faster iteration (no legacy codebase debt)
- Compliance automation that’s native to AI, not bolted on top
Can Intuit out-innovate this startup? Maybe. But history says: probably not. Siebel was bigger than Salesforce in 2000. Siebel couldn’t outrun Salesforce because the problem wasn’t execution—it was that Salesforce’s architecture was designed for SaaS from inception. Siebel’s architecture was optimized for on-prem sales, with all the technical debt that implies.
Intuit has similar risk. Their codebase was built for monolithic tax software. Retrofitting it for AI-native workflows takes time and creates debt. An AI-native startup doesn’t have that problem.
What Actually Matters
The real signal isn’t “does Intuit have AI features?” Everyone will have AI features. The signal is unit economics trajectory. Questions to ask about AI-native competitors:
- How much cheaper is it to acquire customers?
- How much lower are support costs?
- How fast can they iterate on new tax law changes?
- Can they achieve compliance parity in 2-3 years?
If the answer is “yes, and meaningfully better,” then wave three comes for Intuit—probably 2028-2030.
The Actual Winners
Here’s the uncomfortable truth for incumbent investors: the companies that will win wave three probably don’t exist yet. They’ll be founded by someone who sees a gap in the market, builds a purpose-native solution, and gains traction while incumbents are busy defending their installed bases and managing technical debt.
That doesn’t mean Intuit dies or is a bad investment right now. It means the long-term trajectory is pressure—not catastrophe, but margin compression and share loss over a 5-10 year horizon.
The incumbents that survive are either those with genuine regulatory/compliance lock-in so strong that startups can never replicate it (banks, insurance, maybe healthcare), or those rare companies that reinvent early enough and completely enough (Microsoft Azure competing with AWS—and even there, Azure is still losing share because it came late to cloud).
Intuit probably falls somewhere in the middle. Real moats in wave one and two. Pressure in wave three and beyond. Worth owning for the next 2-3 years, not necessarily for the next 10.