Garry Tan SaaS Prediction Scorecard: 3 Months Later (2026)

In December 2025, Y Combinator CEO Garry Tan dropped a prediction that rippled through the tech industry: vibe coding would kill traditional SaaS. He named Zoho as the first casualty. Three months later, $285 billion has been wiped from SaaS valuations, 150,000+ apps have been built on Taskade Genesis alone, and Zoho's CEO fired back with receipts.

Who was right? Let's grade the predictions with data, not opinions. 📊

TL;DR: Garry Tan's SaaS prediction gets a mixed scorecard — 2 confirmed, 2 in progress, 1 too early to call. The $285B SaaSpocalypse and 150K+ vibe-coded apps validate the disruption thesis, but Zoho's 50%+ growth and critical security flaws in generated code show SaaS incumbents are not dead yet. The real winner is workspace-native AI — platforms like Taskade Genesis that combine app generation with agents, automations, and enterprise controls. Try it free →

📋 The Original 5 Predictions

Before we grade anything, here are the five testable claims Garry Tan made in December 2025, sourced from his public statements on X (formerly Twitter):

1. "Zoho's business would be first to be competed away" — a direct shot at the $1B+ revenue SaaS conglomerate
2. "People building their own custom software using @Replit @emergentlabs and @Taskade" — naming specific platforms driving the shift
3. "Why pay $30/seat/month for over-bundled SaaS" — attacking the per-seat pricing model
4. "Even nontech ops people can vibe-code a custom solution in a weekend" — the democratization claim
5. Per-seat SaaS pricing is unsustainable (implied) — the structural thesis underneath it all

Each prediction is testable. Each has three months of market data behind it now. Let's build the scorecard.

📊 Prediction #1: "Zoho Would Be First to Fall"

Garry Tan singled out Zoho by name — a bold move against a company with 100M+ users and decades of profitable growth. Zoho CEO Sridhar Vembu did not take it quietly.

The Evidence

Vembu's counterpunch was immediate and data-backed:

• "If our business would be the first to be competed away, why are we seeing such rapid customer growth (exceeding 50%)?"
• "We will outshine and outlast his vibe coding companies!"
• "Vibe coding just piles up tech debt faster"
• "Why don't we see vibe-coded email or spreadsheet or accounting apps?"

That last question is the sharpest rebuttal. As of March 2026, no one has vibe-coded a production-grade email client, spreadsheet engine, or double-entry accounting system. These are deeply complex domains with decades of edge cases baked into the code.

Zoho's numbers back the talk. The company remains privately held with consistent profitability — no VC burn rate, no IPO pressure, no activist investors demanding AI pivots. Their 50%+ customer growth isn't a vanity metric; it's revenue-correlated adoption in enterprise segments where switching costs are high.

The Verdict: ❌ Refuted

Zoho is growing faster than ever. Tan picked the wrong target. Zoho's integrated suite (CRM + email + accounting + HR + inventory) creates switching costs that vibe coding can't touch. You can vibe-code a dashboard, but you can't vibe-code 15 years of GAAP compliance logic in a weekend.

What Vembu got right: Complex, compliance-heavy SaaS products (accounting, HR, CRM) have deep moats that vibe coding cannot breach in a weekend — or even a quarter. The interconnected data model across 50+ Zoho apps creates lock-in that no single vibe-coded app can replicate.

Why the Zoho Prediction Failed

Tan's mistake was targeting a company that looks like a traditional SaaS vendor but operates like a tech conglomerate. Consider:

• Vertical integration: Zoho builds its own servers, operating system (ZohoOS), and even runs its own data centers in India. That cost structure insulates them from the pricing pressure that hits cloud-dependent SaaS companies.
• SMB stickiness: Zoho's sweet spot is small and medium businesses that adopt 5-10 Zoho apps at once. Ripping out one app means ripping out the data connections to all the others.
• Global distribution: Zoho has massive adoption in India, Southeast Asia, and Africa — markets where vibe coding tooling has lower penetration and where $30/seat is already considered premium.
• No VC clock: As a privately held, profitable company, Zoho faces zero pressure to demonstrate AI-native transformation on a quarterly earnings call. They can adopt AI on their own timeline.

A more accurate version of Tan's thesis would have targeted single-purpose SaaS tools — the one-feature, $20/seat/month products that solve exactly the kind of problem a Genesis app or AI agent workflow can replace in an afternoon. The prediction was right about the category, wrong about the specific company.

📈 Prediction #2: "Platforms Like Replit, Emergent, and Taskade Will Drive the Shift"

Tan named three specific platforms. Three months later, the numbers are in.

The Evidence

The vibe coding market hit $4.7 billion in 2026 with a 38% compound annual growth rate. That's not a niche trend — it's an emerging category.

Taskade Genesis stands out in this landscape because it doesn't just generate code — it generates living applications with built-in AI agents, workflow automations, and 100+ integrations. When 63% of your users are non-developers successfully building and deploying apps, the democratization thesis is more than theory.

Tan himself went beyond predictions — he released gstack, an open-source Claude Code setup that earned 20,000+ GitHub stars and 2,200+ forks. The YC CEO wasn't just talking about the future; he was shipping tools for it.

Platform Quality Divergence

Not all vibe coding platforms are created equal. A critical nuance in the data: the platforms that combine AI generation with workspace infrastructure are outperforming pure code generators on quality and retention.

Taskade Genesis exemplifies this divergence. By embedding app generation inside a workspace that already has AI agents for validation, automations for testing workflows, real-time collaboration for peer review, and persistent Workspace DNA for memory — the quality gap narrows dramatically compared to standalone code generators that dump code into a void.

The lesson: Tan was right that vibe coding platforms would rise. The nuance he didn't articulate is that workspace-native platforms would rise faster and more sustainably than pure code generators.

The Verdict: ✅ Confirmed

Every platform Tan named grew significantly. The broader ecosystem expanded even faster. The evidence is overwhelming: vibe coding platforms are growing, adoption is real, and the market is measured in billions.

💸 Prediction #3: "$30/Seat/Month Is Over-Bundled"

"Why pay $30/seat/month for over-bundled SaaS" — this was Tan's attack on the pricing model that funds most B2B software companies.

The Evidence

The SaaSpocalypse hit in February 2026: $285 billion wiped from SaaS valuations. That's not a correction — it's a repricing of the entire sector.

Key data points that support Tan's thesis:

• Monday.com CEO replaced 100 SDRs with AI agents — if the SaaS companies themselves are replacing human seats with AI, the per-seat model is eating its own tail
• Atlassian reported its first-ever enterprise seat count decline — the bellwether for developer tooling is losing seats
• CNBC analyst independently predicted the same thesis — AI-native wins, SaaS incumbents falter — this wasn't just a VC hot take

Meanwhile, Taskade Genesis offers workspace-native app building starting at $6/month with 11+ frontier AI models, 22+ built-in tools, and 100+ integrations. Compare that to a typical SaaS stack:

The math doesn't lie. When a team of 50 pays $30/seat/month for three SaaS tools, that's $54,000/year. When a single Taskade Genesis workspace can generate custom alternatives with AI agents and automations built in, the CFO starts asking uncomfortable questions.

The Enterprise CFO Calculation

Let's make this concrete. A 200-person company with a typical SaaS stack:

• Project management (Asana/Monday): $25/seat x 200 = $5,000/month
• Help desk (Zendesk): $35/seat x 50 support agents = $1,750/month
• Internal tools (Retool): $40/seat x 30 power users = $1,200/month
• Analytics (Looker/Tableau): $30/seat x 25 analysts = $750/month
• CRM add-ons (various): $20/seat x 80 sales team = $1,600/month

Total: $10,300/month = $123,600/year on layered SaaS subscriptions.

Now compare: a Taskade Business plan at $40/month covers the entire workspace with AI agents, automations, 8 project views, and the ability to generate custom apps for any workflow. Even at the team level, Taskade Pro at $16/month includes 10 users with full Genesis access.

This isn't hypothetical. Teams on the Taskade Community Gallery are already building and sharing custom dashboards, client portals, automated reporting workflows, and AI agent assistants that replace specific SaaS subscriptions.

The $285B valuation wipeout reflects investors doing this same math at scale across the entire SaaS sector.

The Verdict: ✅ Confirmed

$285B in erased valuations, declining seat counts at Atlassian, and SaaS companies themselves replacing seats with AI — the market is repricing $30/seat software in real time.

🛠️ Prediction #4: "Non-Tech Ops People Can Vibe-Code in a Weekend"

This is the democratization claim — the most transformative prediction and the hardest to fully validate.

The Evidence (For)

The adoption numbers are striking:

• Taskade Genesis: 63% of builders are non-developers — operations managers, marketers, founders, consultants building real apps
• Replit: 40M+ users, many from non-technical backgrounds
• Vibe coding market: $4.7B with 38% CAGR — this growth rate requires non-developer adoption at scale
• Andrej Karpathy (who coined "vibe coding" in Feb 2025) declared the concept proven enough to be "passé" by February 2026

Non-developers aren't just experimenting. On Taskade's Community Gallery, you can see real apps built by non-technical users — client portals, project dashboards, automated workflows — running in production.

The Evidence (Against)

Quality concerns are real and well-documented:

The Lovable CVE is particularly damning. When 1 in 10 vibe-coded apps ships with a critical vulnerability, "anyone can build in a weekend" needs an asterisk: anyone can build, but not everyone should deploy without review.

The METR study adds nuance. AI coding tools made developers 19% slower on complex tasks — not because the tools are bad, but because debugging AI-generated code requires different skills than writing code from scratch. For simple CRUD apps, vibe coding saves time. For complex systems, it can create false confidence.

These quality concerns follow a predictable pattern in technology adoption. Early automobiles were less reliable than horses. Early smartphones crashed more than flip phones. The question isn't whether vibe-coded apps have quality issues today — they do — but whether the quality trajectory is improving faster than incumbents can defend. Based on the pace of model improvements, multi-model architectures, and tools like Taskade's built-in AI agents that validate outputs before deployment, the answer appears to be yes.

This is precisely why workspace-native platforms like Taskade Genesis take a different approach. Instead of generating raw code that needs manual review, Genesis creates apps within a workspace ecosystem where AI agents handle validation, automations enforce consistency, and team collaboration provides human oversight.

The Verdict: ⏳ In Progress

Non-technical people are vibe-coding real apps — 63% non-developer adoption on Taskade Genesis proves the concept. But "build in a weekend" and "deploy safely in a weekend" are different claims. The quality gap is closing, not closed.

💀 Prediction #5: Per-Seat SaaS Pricing Is Dying

The structural thesis underneath all of Tan's predictions: the per-seat model is fundamentally incompatible with an AI-native world where agents do the work of humans.

The Evidence

The most compelling data point came from inside SaaS itself. When Monday.com's CEO replaced 100 SDRs with AI agents, he demonstrated that SaaS companies recognize the same disruption Tan predicted — and are cannibalizing their own seat revenue to stay competitive.

Atlassian's first-ever enterprise seat count decline tells the same story from the buyer side. Companies are consolidating tools and reducing per-seat spend.

But the counter-evidence matters too:

• Zoho's 50%+ customer growth suggests bundled SaaS still has massive demand in segments where integration depth > customization freedom
• Enterprise compliance requirements still favor established vendors with SOC 2, HIPAA, and GDPR certifications
• Switching costs remain high for deeply embedded tools (ERP, HCM, accounting)

The market is splitting, not uniformly dying:

The Verdict: ⏳ In Progress

Per-seat pricing is under existential pressure for commodity SaaS. It's not dying overnight for complex enterprise suites. The structural shift is real but segmented — and the speed of disruption depends heavily on which category you're looking at.

🧠 Garry Tan's Evolving Thesis

Tan didn't just make one prediction and walk away. His thinking evolved publicly over the three months since December 2025:

The evolution from hot take to shipped product (gstack) to industry-wide repricing tells a story more nuanced than "SaaS is dead." Tan was right about the direction. He was wrong about the timeline for specific companies like Zoho. And the broader ecosystem moved faster than even he predicted — Karpathy's pivot from "vibe coding" to "agentic engineering" in just 12 months shows how quickly the field is maturing past its initial framing.

The gstack Signal

Tan's decision to release gstack deserves closer analysis. When the CEO of the world's most prestigious startup accelerator releases an open-source developer tool, it sends a specific message: this technology is ready for production use.

The numbers back it up: 20,000+ GitHub stars and 2,200+ forks in under three months means developers aren't just curious — they're building on top of it. gstack's architecture (a Claude Code setup for full-stack development) aligns precisely with the agentic engineering shift Karpathy described. It's not a toy demo — it's infrastructure.

For the scorecard, gstack matters because it shows Tan putting his own effort where his prediction was. Plenty of VCs make predictions on X and move on. Tan predicted vibe coding disruption, then shipped a tool to accelerate it, then watched the market prove him partially right. That's a stronger conviction signal than any tweet.

📉 The SaaSpocalypse: $285B in Context

February 2026's $285 billion SaaS valuation wipeout deserves its own analysis because it's the single largest piece of evidence in this scorecard.

What Drove It

The sell-off wasn't caused by any single event. It was a confluence:

1. AI-native alternatives gaining traction — vibe coding tools, AI agents, and workspace platforms like Taskade demonstrated that custom software was becoming cheaper to build than to buy
2. Enterprise seat count declines — Atlassian's numbers spooked investors across the entire SaaS sector
3. SaaS companies replacing their own employees with AI — Monday.com's 100-SDR replacement showed that AI wasn't just disrupting customers, it was disrupting the sales model
4. CNBC analyst validation — when mainstream financial media independently reaches the same conclusion as the YC CEO, capital follows
5. Broader AI repricing — investors shifting capital from SaaS incumbents to AI-native platforms

What It Means

The $285B wipeout is a market signal, not a death sentence. SaaS revenue didn't drop by $285B — valuations did. That means investors are pricing in lower future growth, higher churn risk, and the possibility that AI-native alternatives will capture share.

For SaaS companies, the response has been predictable: bolt on AI features, launch "AI copilot" products, and rebrand existing automation as "agentic." Whether these defensive moves are enough depends on the category. History suggests that incumbents who add AI as a feature layer on top of legacy architecture lose to AI-native competitors who build from the ground up — just as mobile-first companies beat desktop incumbents who "added a mobile app."

The pattern is already visible. SaaS companies that built AI agents natively into their core product (like Taskade) are growing. SaaS companies that bolted on a ChatGPT wrapper are seeing those features go unused. The $285B repricing reflects investors beginning to distinguish between genuine AI-native architecture and cosmetic AI additions.

For builders and teams evaluating tools, the signal is clearer: the cost of building custom software is plummeting, and platforms that combine AI app generation with workspace intelligence offer more value per dollar than traditional SaaS stacks.

📋 The Complete Scorecard

Here's the full grading of all five predictions, weighted by evidence strength:

Overall grade: 2 Confirmed, 2 In Progress, 1 Refuted.

Tan got the macro direction right. He got the specific target (Zoho) wrong. And the timeline is playing out faster than expected on the market side ($285B wipeout) but slower than predicted on the product quality side (security flaws, data loss incidents).

🔍 What Vembu Got Right

A fair scorecard must credit both sides. Sridhar Vembu made several claims that held up:

1. Zoho's Growth Is Real

50%+ customer growth is not the trajectory of a company being "competed away." Zoho's private ownership gives it the luxury of long-term thinking without quarterly earnings pressure. While public SaaS companies lost $285B in market cap, Zoho kept growing.

2. Vibe Coding Creates Tech Debt

The Lovable CVE-2025-48757 affecting 10.3% of generated apps validates Vembu's "piles up tech debt faster" argument. When AI generates code without deep understanding of security patterns, the tech debt accumulates silently until a vulnerability is discovered.

The Replit Agent database wipe (1,200+ records) demonstrates another form of AI-generated tech debt: operational risk. Code that works in testing can fail catastrophically in production when AI makes assumptions about database operations.

3. Complex SaaS Isn't Vibe-Codeable

Vembu's challenge — "show me a vibe-coded email or accounting app" — remains unanswered. These products require:

• Years of edge-case handling
• Regulatory compliance (GAAP, SOX, HIPAA)
• Deep integration with payment rails, tax authorities, banking APIs
• Battle-tested reliability at scale

No vibe coding platform has produced a production-grade alternative to Zoho Books, QuickBooks, or even Gmail. The gap between "build a dashboard in a weekend" and "build accounting software that passes an audit" is measured in years, not prompts.

📐 Lessons for Tech Predictions

This scorecard reveals patterns useful for evaluating any tech prediction:

1. Direction vs. Target Accuracy

Tan nailed the direction (SaaS faces structural pressure from AI) but missed the target (Zoho specifically). This is common in tech predictions — the macro thesis is easier to get right than the micro execution. "Mobile will disrupt desktop" was right; "Nokia will dominate mobile" was catastrophically wrong.

2. Timing Asymmetry

The financial repricing ($285B wipeout) happened faster than expected. The product quality convergence (vibe-coded apps matching SaaS quality) is happening slower than expected. Markets reprice on narrative before products achieve parity. This creates a window where SaaS valuations are depressed but the actual switching hasn't happened at scale yet — exactly the window we're in now.

3. The Middle Path Isn't Boring

Both Tan (vibe coding kills SaaS) and Vembu (vibe coding is a fad) framed a binary. The reality is a spectrum where workspace-native AI platforms combine the customization of vibe coding with the reliability of enterprise software. The most interesting products aren't at the extremes.

4. Incumbents Underestimate Non-Consumption

Vembu's challenge — "show me a vibe-coded accounting app" — commits the innovator's dilemma error. Disruption rarely attacks incumbents' strongest use cases first. It starts with non-consumption: the internal dashboards, team workflows, and custom tools that companies never built because hiring a developer for a one-off project was too expensive. Those are exactly the 150,000+ Genesis apps being built today.

🏗️ The Workspace-Native Advantage

Both Tan and Vembu miss a crucial third option: the future isn't pure vibe coding or traditional SaaS — it's workspace-native AI that combines both.

Taskade Genesis represents this middle path:

The Workspace DNA model — Memory + Intelligence + Execution — addresses the quality concerns that plague standalone vibe coding tools. When your AI-generated app lives in a workspace where agents validate outputs, automations enforce business rules, and team members provide human oversight, you get the speed of vibe coding with the reliability of structured software.

This is what 150,000+ apps and 63% non-developer adoption looks like in practice: not "anyone can code" but "anyone can build, and the workspace handles the hard parts."

The Build vs. Buy Decision Tree

For teams trying to navigate this transition, the decision framework is straightforward:

Build with Taskade Genesis when:

• The tool is team-specific (custom dashboard, internal tracker, client portal)
• You need AI agents embedded in the workflow, not bolted on
• You want to iterate quickly without waiting for a vendor's product roadmap
• The cost of SaaS seats exceeds $6/month per user
• You need 100+ integrations connecting to your existing tools

Keep existing SaaS when:

• The product handles regulated data (healthcare records, financial audits)
• Your entire team's workflows depend on the vendor's specific data model
• Switching costs exceed 6 months of SaaS spend
• The vendor offers genuine AI-native features (not just a chatbot wrapper)

Most teams will find that 30-50% of their SaaS stack falls into the "build" category — exactly the commodity tools that Tan predicted would be disrupted first.

🔮 What Comes Next: From Vibe Coding to Agentic Engineering

Andrej Karpathy's February 2026 shift from "vibe coding" to "agentic engineering" signals where this entire category is heading.

Vibe coding (2025): Describe what you want, AI writes the code. Simple, accessible, but quality-limited.

Agentic engineering (2026+): AI agents don't just write code — they plan, test, deploy, monitor, and iterate. The human provides intent and judgment; the agents handle execution across the full software lifecycle.

This evolution explains why Taskade Genesis invested early in AI agents with persistent memory, workflow automations with 100+ integrations, and workspace-native intelligence. The vibe coding wave was chapter one. Agentic engineering is chapter two. And the platforms that combine both with team collaboration will capture the most value.

What Agentic Engineering Means in Practice

The shift from vibe coding to agentic engineering isn't just a label change. It represents a fundamental evolution in how AI-generated software is built, tested, and maintained:

1. Planning agents analyze requirements and break projects into structured tasks — not just "write code for X" but "design the architecture, identify edge cases, plan the test suite"
2. Building agents generate code with context awareness — understanding the full codebase, not just the current prompt
3. Testing agents automatically verify output against requirements — catching the kind of security flaws that created the Lovable CVE
4. Monitoring agents watch deployed apps for performance degradation, security vulnerabilities, and user experience issues
5. Iteration agents respond to feedback loops — user reports a bug, agent diagnoses and fixes it autonomously

Taskade Genesis already implements elements of this pipeline. When you build an app, AI agents with 22+ built-in tools can validate business logic, automations handle deployment workflows, and the Workspace DNA system ensures persistent memory across the entire lifecycle. It's not just code generation — it's the full Memory + Intelligence + Execution loop.

This is why the distinction matters for Tan's predictions: he was describing a vibe coding future, but the market is already moving past it into something more robust. The platforms that win will be the ones that made the agentic engineering leap early.

📊 Market Context: Where the Money Is Moving

The broader market data adds context to the scorecard:

Capital is flowing into AI-native platforms and out of traditional SaaS. The question isn't whether the shift is happening — it's how fast, and which platforms capture the value on the other side.

👥 What This Means for Different Stakeholders

The scorecard has different implications depending on who you are:

For SaaS Founders

Your moat is no longer features — it's data gravity and workflow integration. Single-feature tools charging $20+/seat are the most vulnerable. If your entire value proposition can be described in one sentence ("we help teams track X"), a Genesis app or AI agent can likely replicate it. The playbook: go deeper into compliance, add AI-native capabilities, or accept that your product will be commoditized.

For Enterprise Buyers

You now have legitimate alternatives to the SaaS stack. The Taskade Genesis model — build custom tools within a workspace that already has AI agents, automations, and team collaboration — eliminates vendor lock-in for commodity workflows. Start by identifying your simplest SaaS subscriptions and testing whether a custom Genesis app or automated workflow can replace them. The math usually works at 3+ seats.

For Non-Technical Builders

The democratization is real but the quality bar matters. Don't vibe-code a payment processing system without professional review. Do vibe-code internal dashboards, project trackers, client portals, reporting tools, and team workflows. Taskade Genesis is specifically designed for this use case — the workspace handles security, collaboration, and infrastructure so you can focus on the business logic.

For Developers

Your role is shifting from "person who writes code" to "person who architects systems and reviews AI output." The METR study's 19% slowdown finding actually supports your value — complex systems still need human oversight. The opportunity is in agentic engineering: designing the agent workflows, setting up the guardrails, and building the custom tools that make AI-generated apps production-ready.

For Investors

The $285B wipeout is a repricing, not a collapse. SaaS revenue is still flowing — it's the growth premium that evaporated. Smart money is moving into platforms that combine AI-native capabilities with workspace infrastructure. The winners will be companies like Taskade that serve both the builder (Genesis) and the team (workspace) — not pure code generators with no retention model.

⚖️ The Real Verdict

Garry Tan's December 2025 predictions were directionally right but specifically wrong.

What he got right:

• SaaS valuations would face a reckoning ($285B wiped)
• Vibe coding platforms would grow massively ($4.7B market)
• Non-technical people would build real software (63% non-developer adoption)
• The per-seat pricing model would come under pressure (Atlassian declining, Monday.com replacing seats)

What he got wrong:

• Zoho would be the first casualty (Zoho is growing 50%+)
• The timeline for quality parity (security flaws, data loss incidents show the gap is real)
• The binary framing (vibe coding or SaaS) — the real answer is workspace-native AI that combines both

What Vembu got right:

• Zoho's growth would continue (50%+ confirmed)
• Vibe coding creates tech debt (Lovable CVE, Replit data loss)
• Complex SaaS cannot be weekend-coded (no production accounting or email apps exist)

What Vembu got wrong:

• Dismissing the structural threat (the $285B SaaSpocalypse is not a blip)
• Assuming vibe coding is a fad (it's a $4.7B market growing at 38% CAGR)
• Ignoring the middle ground — workspace-native platforms like Taskade Genesis that bridge the gap

The future isn't SaaS versus vibe coding. It's platforms that combine the best of both: the customization and speed of AI-generated apps with the reliability, team collaboration, and enterprise controls of mature workspace software.

That's exactly what Taskade Genesis delivers — 150,000+ apps built, 63% non-developer users, 11+ frontier AI models, 22+ agent tools, 100+ integrations, and Workspace DNA that turns every prompt into a living, intelligent system.

🔭 What to Watch in Q2 2026

The scorecard will keep evolving. Here are the key metrics and events to track over the next three months:

Metrics That Will Shift the Scorecard

• Zoho's next growth report — if Vembu's 50%+ holds through Q2, Prediction #1 stays firmly refuted. Any deceleration reopens the debate.
• Vibe coding security incidents — more CVEs like the Lovable vulnerability will slow enterprise adoption. Fewer incidents will accelerate the quality narrative.
• Atlassian Q2 seat counts — a second consecutive quarter of enterprise seat declines would confirm a structural trend, not a one-time blip.
• SaaS valuation recovery — if the $285B wipeout was a temporary fear trade, valuations should partially recover. If they stay depressed, the market is pricing in permanent disruption.
• Taskade Genesis app count trajectory — the 150,000+ milestone is the baseline. The growth rate from here signals whether non-developer adoption is accelerating or plateauing.

Events to Watch

• Y Combinator W2026 Demo Day — Tan's portfolio will reveal how many YC companies are building AI-native SaaS replacements
• Google I/O and WWDC — if Apple or Google introduces native vibe coding capabilities in their platforms, the market expands massively
• Enterprise earning season (Q1 reports) — SaaS companies will disclose AI impact on retention and expansion revenue for the first time
• Regulatory developments — EU AI Act enforcement begins affecting AI-generated software, potentially creating compliance moats for established platforms with proper guardrails

The Tan vs. Vembu debate isn't over. It's entering its most decisive phase.

🚀 Build the Future — Don't Wait for the Verdict

The SaaS disruption debate will continue for years. You don't have to wait for a winner to start building.

Try Taskade Genesis free →

Build your first app in 30 seconds. No code required. Your workspace handles the rest — AI agents for intelligence, automations for execution, team views for collaboration.

• 150,000+ apps built and counting
• 63% of builders are non-developers
• Starting at $6/month (annual billing)
• 100+ integrations across 10 categories
• Community Gallery with thousands of real apps to clone and customize

Whether Garry Tan or Sridhar Vembu is ultimately proven right, the teams building with AI-native workspace tools today will be the ones best positioned tomorrow.

Start building → | Explore the Community → | See pricing →

Frequently Asked Questions