Klarna CEO: SaaS is Dead: Why Systems of Record Will Die in an Agentic World

TL;DR

Overall Summary: AI agents will collapse software switching costs by enabling one-click data migration, destroying the moats that let SaaS companies charge premium multiples.

Key Points

• 1.SaaS moats are crumbling: Software creation cost is approaching zero; the next collapse is switching costs as AI agents migrate data between systems effortlessly
• 2.Valuations will crater: SaaS companies trading at 5-10x price-to-sales could fall to 1-2x (utility-level multiples) as switching costs vanish
• 3.Klarna killed internal SaaS: They consolidated data from siloed SaaS tools into one AI-native system to give their AI better context for decision-making
• 4.AI needs unified data: Fragmented data across CRMs, ERPs, and project tools cripples AI performance—companies must consolidate or lose
• 5.Customer service is a test case: Klarna's AI replaced 700 agents' worth of work; the AI reads source code directly to answer customer questions accurately
• 6.Enterprise software builders will vibe-code: Large companies will build custom internal tools; small companies will buy "company in a box" AI solutions
• 7.Headcount collapse is real: Klarna went from 7,000 to under 3,000 employees (50% reduction) through attrition, projecting potentially below 2,000 by 2030
• 8.Human connection becomes premium: VIP service will mean talking to actual humans; Klarna now recruits passionate customers as part-time support via Uber-style model
• 9.Per-seat pricing is dead: Investors should only back companies that displace labor, not charge per seat
• 10.Incumbent banks face extinction: Tech-forward fintechs (Klarna, Revolut, Nubank) are eating market share; banks that don't reinvent will fade
• 11.The future is agents assembling Lego blocks: AI won't code from scratch—it'll stitch together cached, standardized, security-assured components