$Duolingo(DUOL.US) From a certain perspective, DUOL is even in a safer position than Microsoft 🤔

The impact of AI on 2C SaaS is indeed not as significant as on 2B SaaS—especially in terms of disruptive/negative effects. 2B faces clearly greater risks and pressure. Many analyses even call this the "Great SaaS Divide" or "B2B SaaS being killed by AI."

Why is 2B more impacted?

B2B SaaS (like Salesforce, ServiceNow, Workday, Adobe, etc.) are essentially utility tools, selling ROI, efficiency, and compliance. Enterprise clients are rational decision-makers:

• If AI (or "Model-as-a-Service") can do the same job at an extremely low cost, enterprises will switch even if migration is troublesome—because saving millions of dollars is too tempting.

• After AI improves efficiency, it directly leads to a reduction in seats: one person can do the work of several before, so corporate procurement departments will cut subscriptions, pressure prices, and shift to outcome-based pricing.

• Result: Revenue growth rates for traditional B2B giants have significantly slowed (Salesforce dropped from 24.7% in 2022 to a projected 8.4% in 2025, similar for ServiceNow and Workday). Valuation pressure forces many companies to transform, otherwise the "middleware" layer gets eaten by AI.

Infrastructure exceptions (like CrowdStrike, Datadog): AI actually increases demand for security and data, making them more stable.

Why is the 2C moat stronger, with less impact?

B2C SaaS (like Spotify, Netflix, Duolingo, Notion, Canva, etc.) sell emotion, habit, and identity:

• Users aren't using them for "efficiency," but for daily rituals (Duolingo streaks, Spotify background music, Netflix relaxing binge-watching). AI finds it hard to break this "why bother" emotional stickiness.

• Low pricing (a few to over 20 dollars per month), users can't be bothered to switch to save money; can also monetize through ads, IP, and content diversification.

• AI in 2C is more of an enhancer: helping generate content, personalize recommendations, lower creation barriers, actually expanding the user base and increasing stickiness, not replacing the entire platform.

In a nutshell: B2B is a math problem (rational calculation), B2C is a relationship problem (emotional retention). AI is good at solving math problems, but not human habits and pleasure.

Other observations (2025-2026 data trends)

• The Agentic AI era: B2C/bottom-up products benefit from "everyone can use," expanding user base; enterprise B2B may see revenue stagnate or even decline due to headcount reduction and budget squeeze.

• Enterprise AI budgets surge (+100%+), but total IT budgets only increase slightly, with money directly squeezed out of traditional SaaS.

• Mistral's CEO even said: Over 50% of enterprise SaaS will eventually be replaced by AI custom applications—this basically refers to 2B.

Of course, both have opportunities: 2B needs to quickly shift to consumption-based pricing, infrastructure, AI agent integration; 2C should use AI for better personalization, content generation, and defending daily rituals.