Howie Liu - Building Airtable - [Invest Like the Best, EP.375]

Episode Summary

Executive Summary: Patrick O'Shaughnessy interviews Airtable CEO Howie Liu on how AI changes software. Liu argues Airtable’s platform, distribution, and workflow/data model make it well positioned to turn LLMs into useful business applications, especially where broad models need human-in-the-loop workflows and customization.

Main Topics: Airtable’s AI strategy (Priority: 5/5): Use best models as infrastructure, then apply them to custom workflows and customer data. Horizontal but deep software (Priority: 5/5): Liu argues the winning category is platform software that is broadly applicable and highly customizable. AI as workflow transformation (Priority: 5/5): He prefers end-to-end process redesign over shallow AI features like summaries. Enterprise adoption constraints (Priority: 4/5): He thinks enterprise uptake will be slower because behavior and process change lag model capability. Airtable’s evolution as a product (Priority: 4/5): The company moved from blank-slate Lego blocks to templates, blueprints, and solution layers. Crucible moments in Airtable’s history (Priority: 4/5): He traces key inflections from product-market fit to monetization, scale, COVID, and efficiency. What future winners will look like (Priority: 4/5): He expects AI-first industry operators and platform players to reshape major sectors.

Key Arguments: Airtable wins by combining distribution, data, and workflow layers that LLMs need. Horizontal platforms can be deep when users assemble custom apps from building blocks. AI value comes from process transformation, not isolated features like thread summaries. Enterprise adoption will be slower than SMB/consumer because behavior change is hard. Current LLMs already create trillions in potential value if embedded in real workflows. Better reasoning, consistency, and easier prompting matter more than flashy novelty. Airtable’s role is to help customers design, not just consume, AI-enabled workflows.

Data Points: organizations using Airtable: over 300,000 organizations - Howie describes Airtable’s installed base and reach paying customers: around 100,000 total paid customers - Includes SMB self-serve customers Fortune 500 penetration: more than half the Fortune 500 - He cites enterprise distribution as a key advantage founded: 2012 - Airtable founding year pre-launch build period: the first three years - He says the team spent this long building before launch first monetization milestone: $10,000 customer - Early sign that users would pay for Airtable rapid revenue growth: 500k to 1 million in a few months - Describes early monetization acceleration scale milestone: 10 million to 20 million in less than another year - Shows the business scaling quickly headcount growth: more than 100 per year - Company scale-up during the growth years workforce size: low 100s to over 1,000 - Headcount increase over a few years revenue growth: 40 plus percent revenue growth rate - Airtable’s overall growth after the market shift runway at founding: two and a half months of runway on a ramen budget - Liu describes his personal early startup constraints YC funding: $15,000 - Amount provided by Y Combinator at the time

Pivotal Quotes: "we actually give them all the building blocks they need to completely customize the use case they have." — Howie Liu: Explaining Airtable’s approach to AI and platform software "the big gap is really around having that data and workflows and human interface layer to them." — Howie Liu: Why LLMs need a platform to create value in enterprises "I think the most special advancement, in my opinion, is going to be the quality of the reasoning." — Howie Liu: What matters most in future model improvements

Implications: The open question is which workflows will prove durable, so teams should focus on high-friction processes where AI can be embedded end to end.