The Only SaaS Moats Left in 2026

gist

In a MicroConf talk co-written with Einar Vollset, Rob Walling lists ten ways AI attacks a SaaS business, six on supply and four on demand, and guesses that 10 to 20 percent of SaaS companies will decline or go to zero. He argues that unique features, owned traffic, clever code and UX familiarity no longer protect a product. He then names eight moats that still work, from trusted buyers and systems of record to regulatory approval, network effects, learning loops and software that is hard to keep running.

ideas

• Supply shocks flood the market. Customers rebuild the three features they use, competitors reach parity fast (one slide shows 8 of 46 AI note takers), open source gets deployable, platforms ship your feature, agencies sell custom builds at SaaS prices, and models absorb products.
• Demand shocks shrink the buyers. Walling warns that AI can wipe out customers such as translation agencies, cut support seats from about eight humans to one or two supervisors, and remove the human user entirely.
• Acting beats advising. He contrasts tools that recommend a payment with ones that make it, citing Stripe Connect, Reimbi and Churnkey, and pairs this with outputs nobody checks, such as Gusto payroll and Avalara sales tax.
• Approval and coordination are slow to copy. Certifications, board-approved ABA courses and SMS carrier registration take months, and products like Slack or a swim-league app force a whole community to move at once.
• Data counts only as a learning loop. Stripe Radar qualifies because every merchant improves fraud detection for the rest, which passes his three gates: shared improvement, learning that can't be exported, and continued learning.

quotes