Howie Liu is the co-founder and CEO of Airtable. We cover the practical applications of AI for making workflow more productive, the most important variables for software investors to consider, and some of the difficult challenges he's faced while building Airtable.
Principles & Lessons:
1) The role of “horizontal yet deep” platforms in capturing changing demands. Howie stressed that “we have a product that basically is an app platform” and not just a simple, shallow productivity tool, arguing that “you can build a workflow layer on top of that” to handle very specific use cases. He contrasted this with more traditional “verticalized solutions” that solve just one industry problem, or with “broad but shallow” products like Office. He said that “the reason we have companies like Netflix early on building these content production systems” is precisely because Airtable offers flexible components—like “LEGO pieces”—that a customer can adapt to their specialized workflow.
2) The indispensable link between data, workflow, and AI. Howie explained, “we can integrate with existing systems of record, you can build a workflow layer on top of that, and it turns out those are all the ingredients that I think are completely necessary to actually unlock value out of these LLMs.” He expressed skepticism about “a pure, agentic LLM experience that just does a bunch of stuff behind the scenes,” suggesting that AI needs a structured environment where humans can model data, review outputs, and guide the process rather than expect the model to handle everything autonomously.
3) True adoption of AI requires deeper “process transformation,” not single-feature additions. Howie said he wants “products that are not just AI as a feature.” He noted, “I turned on Slack’s AI capabilities… you can summarize a thread… but that’s a lot less strategically important to us,” emphasizing that real value emerges from rethinking “the actual job to be done, the business process, the use case.” For example, “helping with the generation of product requirements documents, creating a beautifully summarized e-mail or weekly updates of what’s going on in product” are tangible changes to how teams collaborate.
4) Customizability will spread because of easier building—and easier usage—powered by AI. Pointing to Salesforce’s success, Howie said, “Salesforce wins… not because they went and built the best possible CRM… but rather that they built a platform.” He expects that, as AI design and generation improve, “AI will for sure accelerate” the movement toward every organization—large or small—wanting a high level of flexibility. “If I work at Nike… maybe the AI can infer exactly how any given person within Nike wants to build out their use case and do it for them.”
5) The decisive importance of consistent, high-quality LLM outputs. Howie highlighted, “the most special advancement… is going to be the quality of the reasoning.” He noted that chaining together tasks can cause drift, so “we may look back in a few years and realize that we had all the pieces of a much more advanced form of intelligence… we just didn’t quite figure out how to chain them together.” Improving stability and reliability of results—even from 80% correct to 95-99%—makes “a top of the class” assistant that people will trust to handle real business processes.
6) “Crucible moments” in building a business often reflect shifts in external markets and internal priorities. Howie recounted how COVID triggered panic and immediate churn from some SMB customers: “We were worried because a lot of our customers were SMBs… some were literally shutting down their own business.” But the sudden push to hybrid or remote work ended up providing a major “inflection in growth” for Airtable, as they offered “a modern, much faster way to stand up these 2.0 operations.” He also spoke of repeated decisions to refine headcount and maintain agility, even when capital was ample.
7) Major leaps in usage happen when customers go from novelty to real willingness to pay. Howie recalled that, once they introduced monetization, “we had $500,000 in revenue… then $1 million… then $10 million… it was a really rapid acceleration.” He described the “aha moment” of landing their first $10,000 customer: “I still feel this immense sense of awe… it’s pure software… a service that lives in the cloud.” That early proof of genuine business value—rather than just user sign-ups—was pivotal in moving the product from a beloved indie tool into a scalable company.
8) Small, empowered teams can innovate faster than huge resource pools. Discussing why he made tough decisions like two separate RIFs, Howie stated, “I actually think you need smaller, tighter-knit teams that work more directly with customers to really understand… the jobs to be done and figure out how to actually make it real very, very quickly.” He referenced Apple’s Macintosh as “made by 50 people” who moved like “a renegade team,” underscoring the belief that “it’s not a problem that throwing thousands of people at it will solve faster.”
Transcript