General Catalyst, per partner Zeynep Yavuz-Willson, is an investment company that is also an AI transformation firm focused on early-stage bets.
Zeynep Yavuz-Willson says fragmented service markets like property management and real estate resist AI adoption because buyers do not follow each other, unlike law where firms copy one another.
Zeynep Yavuz-Willson argues customer trust and engagement are now AI's main defensibility, and that applied platforms like Legora can use their own data to improve their models.
General Catalyst bets AI's biggest prize lies in fragmented service industries
General Catalyst argues the largest AI opportunity lies in service industries such as accounting, property management, insurance and law, which are far bigger than software markets. Some buy technology readily, while fragmented ones do not, so the firm funds roll-ups and treats them as customer acquisition. In both cases the lasting advantage is customer trust and the data that applied platforms generate.
Takeaways
Service categories are the bigger AI prize than software. General Catalyst mapped where AI would have the most impact and chose accounting, property management, real estate, law, wealth and insurance. Zeynep says these markets dwarf software in size and touch everyday life, which is why the firm calls itself an AI transformation firm.
Buying behavior splits industries into two kinds. Law buys technology readily because big firms copy each other, and Legora benefits from that pull. Property management, real estate agency, accounting and insurance are fragmented, so buyers do not follow each other and market pull is absent, especially in Europe.
Fragmented markets need roll-ups, which count as customer acquisition. Dwelly's founders saw that organic growth is structurally hard in UK lettings, where contracts run three to five years. The firm backed inorganic growth through roll-up strategies and treats the acquisition spend as customer acquisition cost, not a venture mandate problem.
Customer trust and data are the main AI moats. Zeynep sees the customer relationship and trust as the core defensibility, citing Legora's engagement and word-of-mouth at Langdock. Data from platforms like Legora, Dwelly and Integral can become a reinforcement learning environment that improves their own models.
Open source favors application companies, despite rising costs. Zeynep calls open source a tailwind for the applied layer, even as AI costs have risen since December and frontier labs like Anthropic move into applications. Sebastian says apps may swap out model providers, and she agrees the engagement layer should win.
Accounting adoption depends on partnering with incumbents. Germany's DATEV holds roughly 100% share and is owned by accounting firms, so startups must integrate and build automation on top. Edge cases and zero tolerance for filing errors make it harder. Concentrated markets adopt faster through boardroom decisions.
Capital and incubation tools extend beyond standard venture. CVF funds go-to-market spend without dilution, so equity can cover risky product bets, as with Factorial. EU AI Champions pairs firms like Heineken and Miele with AI startups. The firm also incubated Percepta and Hatch and bought a US health system.
Tap a name to go there · tap a line to see why they are linked