/
.webp)
Generated by Master Biographer | Source for LinkedIn Content
San Francisco. Sometime in 2012.
Howie Liu is watching a coworker do something ordinary. The kind of thing that happens in every office in every city in the world, so common that nobody ever bothers to question it. The coworker has a spreadsheet open. Columns for names. Columns for dates. A column called "Status" with hand-keyed values — "In Progress," "Done," "Waiting on someone else." Color-coded rows, each shade applied by hand. A frozen top row. Filter dropdowns.
They're not doing math. There's not a formula in the file.
They are, very carefully and very manually, building a database. Without knowing it.
Liu had been at Salesforce for about a year at this point. He'd come in through an acquisition — his first startup, Etacts, a contact management tool he'd built at 21 with his co-founder Evan Beard, had been scooped up by Salesforce in December 2010 after catching the attention of investors including Ashton Kutcher at a Y Combinator Demo Day. Liu was handed a product manager role leading the social CRM team. Good title. Smart colleagues. Access to the machinery of one of the most powerful enterprise software companies in the world.
And still, this spreadsheet.
The thing that lodged itself in his mind was not the spreadsheet itself. It was the implication. He started pulling the thread. If this person was doing this, how many others were too? He found the answer, and it was a round, stunning number:
Almost 90% of spreadsheets don't have formulas.
The tool that Lotus 1-2-3 and VisiCalc and Microsoft Excel had spent decades perfecting — the tool of the financial analyst, the accountant, the quantitative modeler — was being used by the vast majority of its users for something else entirely. Not computation. Organization. People were bending a calculation engine into a record-keeping system. They were pouring relational data into flat cells. They were reinventing, by hand, every day, something that database technology solved in the 1970s.
They were using a spreadsheet because a real database was impossible for them.
The CRM they needed was too expensive. The custom app would take months and a developer they didn't have. The project management software was too rigid, designed for someone else's workflow. So they opened a spreadsheet and started typing.
Liu had spent the last eighteen months inside Salesforce watching the company do something brilliant: take the complexity of enterprise software and distill it into a customizable platform. Every Salesforce installation was, at its core, a version of the same underlying data model. Objects. Relationships. Fields. Views on top. The power was in the abstraction — you didn't need to rebuild the architecture each time, you just configured it.
What Salesforce had done for large enterprises with large IT teams, nobody had done for everyone else.
The coworker was still typing status updates by hand.
Howie Liu graduated from Duke University in 2009 with a degree in mechanical engineering and public policy — an unusual combination that already signaled something about how his mind worked. Not purely technical. Not purely strategic. Somewhere in between, at the intersection of systems and people.
He was 20 when he cofounded Etacts. Twenty. The company built a CRM product for managing contacts through email, essentially a smarter address book with memory. It wasn't transformational, but it worked, and it attracted attention. When Salesforce acquired it a year later, Liu was 21 years old with one exit behind him, a product manager title in front of him, and already a growing conviction that the acquisition was not the destination.
At Salesforce, he reported through a chain that ended at Alex Dayon, the company's president of products. He watched how real enterprise software was bought, configured, and deployed. He saw what it cost — not just in dollars but in time, in headcount, in months of implementation before a single user got value. He saw the consultants, the admin certifications, the multi-year contracts.
And he kept noticing the spreadsheets.
Not just at Salesforce. Everywhere. People building CRMs in Google Sheets. Event planners tracking RSVPs in Excel. Nonprofit program managers running donor databases in rows and columns. Video producers logging cast and crew in a flat grid. Every one of them constructing, brick by laborious brick, a broken version of something a relational database could have given them instantly — if only they could use one.
Liu's insight was precise: the problem wasn't that people were bad at spreadsheets. The problem was that people had a need — structured, queryable, relational data — and the only tool accessible to them was a spreadsheet. Nobody had built the bridge between those two things. Salesforce sat on one side, powerful and expensive and requiring an army of specialists. Excel sat on the other, familiar and immediate and fundamentally limited.
The bridge didn't exist.
He left Salesforce in early 2013. He'd been there about a year and a half. He was 23.
Two people came with him.
Andrew Ofstad had graduated from Duke two years before Liu. After college he went to Accenture's research and development lab building prototypes, then moved to Google, where he worked on Google Maps. He was a designer's designer — obsessive about interaction, about what it felt like to touch a piece of software. His intuition about user experience was the kind you can't teach. He would be responsible for making the product feel right, which in the case of Airtable was everything. A spreadsheet-database hybrid that was hard to use was worthless. It had to be instantly comprehensible. It had to feel like something people already knew.
Emmett Nicholas had also gone through Duke's orbit, then moved to Palantir as an engineer. He would handle the architecture. Building a product that was simultaneously as flexible as a spreadsheet and as structurally sound as a relational database was a genuinely hard engineering problem. You had to give users the freedom to define their own schema without letting them break the underlying data integrity. You had to allow linked records across tables without forcing users to understand foreign keys. Every technical decision had to be invisible to the person sitting at the keyboard.
The three of them — the product visionary, the experience designer, the systems architect — took a different approach to building than was fashionable in Silicon Valley in 2013.
The lean startup methodology said: launch early, fail fast, iterate. Put something broken in front of users and learn.
They said: no.
"We took more of a first principles approach," Ofstad later recalled. "We didn't want to just throw spit at the wall."
They spent months studying the history of computing. Douglas Engelbart. Xerox PARC. The origin of the graphical user interface. The principle they kept returning to was not "minimum viable product" but something closer to "maximum comprehensibility." The GUI had made computing accessible to everyone by abstracting away the command line. What Airtable needed to do was abstract away the database without making it feel dumbed down.
The first prototype didn't persist any data. It was purely client-side — a way to test whether the interactions made sense. Whether a person with no database knowledge could look at the interface and understand that they were building something structured. That a cell could be more than a cell. That a column could have a type — date, attachment, checkbox, linked record — and that the software would enforce that type so the user didn't have to.
They kept the feedback loop extremely tight. About 100 private users. Ofstad later noted that adding more engineers would have been "counterproductive" — the work was so interdependent, so reliant on coherent design decisions propagating through the entire product, that parallelizing it would have introduced more confusion than speed.
They spent eight months on the iPhone app before they'd even launched publicly. This was, Ofstad would later admit, a mistake. They'd assumed Airtable would spread the way Evernote and Dropbox did — through personal use, individual adoption, the phone as the primary touch point. They were wrong. Airtable would turn out to be a team product, a workplace product. The mobile investment was real work in the wrong direction.
These are the kinds of mistakes you only recognize years later, when the shape of the thing becomes clear.
By 2014, they had something to show people. Not everything. Not the finished product. But enough.
They launched an invite-only beta. Posted on Hacker News. Positioned the product as — and this phrase, functional and honest and slightly awkward, would follow them for years — "a real-time spreadsheet-database hybrid."
It was hard to explain. This was the core challenge of building a horizontal product: when you can be used for almost anything, it's almost impossible to say what you are. Are you a database? Are you a spreadsheet? Are you a project management tool? Yes, yes, and yes — which was an answer that made investors nervous and marketers miserable.
About one in ten investors they pitched got excited. The other nine heard "spreadsheet" and saw a crowded commodity market. They heard "database" and thought of enterprise software with long sales cycles. The combination didn't click for them.
The founders kept building.
March 2015: public launch. A $3 million seed round led by Caffeinated Capital. Then a $7.6 million Series A in May, led by CRV. The product was live, growing — but slowly.
The first users who latched onto Airtable were not who the team had expected. They'd assumed the earliest adopters would be data analysts, people who understood relational databases and were frustrated by their limitations. Instead, what they got were "systems thinkers" — people in non-technical roles who had a pathological need to organize information and an intuition for structure that never found a tool worthy of it.
Video producers. Event planners. Nonprofit program managers. Content teams building editorial calendars. Marketing operations managers building campaign trackers. People who had been wrestling spreadsheets into submission for years and suddenly found a tool that bent to their will instead of the other way around.
These early adopters did something crucial: they pulled in their colleagues. One user inside a company became a team. A team became a department. The product spread laterally, person to person, the way the best software does — through direct experience of it working, not through a sales pitch.
Airtable tracked this obsessively. They built network graphs of how the product traveled through organizations. They identified what they called "golden datasets" — the specific use cases that drove viral adoption: product roadmaps, marketing calendars, content pipelines. When someone built one of these in Airtable and shared it with their team, the team adopted the tool. Then the tool spread beyond the team.
The company grew. But slowly. This was not the hockey stick of a consumer app finding its audience. It was something steadier, more structural — the gradual replacement of duct-taped spreadsheet systems with something that actually worked.
Liu had told the team early: structure for the long game. The company raised enough capital to maintain eight years of runway. It did not rush. It did not pivot. It did not chase a niche because a niche was faster to win.
This patience cost them visibility. From 2013 to 2016, Airtable was largely a product insiders knew and loved and couldn't quite explain to their friends. The "no-code" movement had not yet acquired its name, its manifesto, or its venture capital. The idea that non-technical people deserved software tools as powerful as those built by engineers was a conviction, not yet a category.
Airtable was ahead of the category by three years.
Being early is not the same as being wrong. It is, however, often indistinguishable from being wrong, until suddenly it isn't.
The moment Andrew Ofstad understood that Airtable had become something real did not happen on a dashboard. It did not happen in a board meeting. It happened in a hallway.
He was visiting WeWork's offices in New York. A customer tour, the kind of thing you do when you're trying to understand how people use your product in the wild. He walked through the open floor plan. Desks arranged in clusters. Laptops open.
He looked around.
Every computer monitor had Airtable open.
Not a demo. Not a pilot. Just: everyone using it. Using it the way you use something that has become part of how you work, something you reach for without thinking about it.
"Oh my God," Ofstad thought. "This is actually a thing."
He later described it as the moment the product became real for him in a way it hadn't been from inside the building. All the metrics, all the user conversations, all the support tickets and NPS scores — none of it had quite felt like this. Looking around a real office and seeing your software on every screen.
"It became so much more real at that point."
By 2016, WeWork was not the only story like this. 2 million bases had been created on the platform. The product was spreading through media companies, marketing agencies, research organizations, film and television productions, nonprofits. Each vertical had found its own way in. Each had bent the tool to its own specific shape.
That was the point. The tool was meant to bend.
November 2016: Airtable launched kanban boards. The same underlying data, viewable as cards moving through stages. The same table — now a Trello board, essentially, but connected to everything else in your Airtable base. Suddenly the project management use case clicked for tens of thousands of users who hadn't thought of Airtable as a project management tool before.
Each new view was not a new product. It was a new lens on the same data. The elegance of this architecture — one structure, infinite presentations — was the thing Liu had been building toward since he watched that coworker color-code their spreadsheet by hand.
The money started coming faster than the company could plan for it.
2018: a $52 million Series B, with the launch of Airtable Blocks — customizable extensions that let users embed charts, maps, timelines, and third-party integrations directly inside their bases. Suddenly the product wasn't just a database with views; it was an app-building platform. Liu said at the time: custom apps that would have cost hundreds of thousands of dollars and months of engineering time could now be built in days by the people who actually needed them.
That year, Airtable claimed 30,000 business users, 500% year-over-year revenue growth, and clients including Airbnb, Condé Nast Entertainment, Penguin Random House, and Tesla. The Tesla contract was reported to be in the six figures.
November 2018: $100 million Series C, led by Benchmark and Thrive Capital. The unicorn label arrived. The $1.1 billion valuation. Liu was 28.
2020: $185 million Series D at a $2.5 billion valuation. The pandemic had done something unexpected for Airtable — remote work had accelerated exactly the kind of adoption it was built for. Suddenly teams scattered across cities and time zones needed flexible, shareable operational infrastructure. The product that had been slowly infiltrating organizations for six years was now the obvious answer to a newly urgent question.
The no-code movement had a name now. Venture capital had discovered it. Conferences were held in its honor. "Citizen developer" became a phrase people said without embarrassment. The world had caught up to the thing Airtable had been building since 2013.
2021: two more rounds in the same year. A $270 million Series E. Then, in December, the capstone.
$735 million. Series F. Investors including Silver Lake, Franklin Templeton, J.P. Morgan Growth Equity Partners, MSD Capital (Michael Dell's family office), Salesforce Ventures — the company that had started it all, as an acquirer of Etacts a decade earlier, was now a capital partner in Airtable. The valuation: $11 billion.
"As we've sharpened our focus on the needs of enterprise businesses," Liu said in the announcement, "customers are scaling their use of our products to department- and company-wide processes, establishing Airtable as a source of truth for their most valuable operations."
300,000 organizations were on the platform. 80% of the Fortune 100 used it somewhere. The 23-year-old who left Salesforce with a conviction that spreadsheets were being misused had built, in eight years, one of the most valuable software companies in the world.
Then came 2022.
Howie Liu had said "this additional capital will allow us to invest even more aggressively." And they had. The team had grown at breakneck pace. They'd hired for engineering, for sales, for product, for business development. The enterprise pivot required a different kind of organization — sales teams, customer success, enterprise implementation. The bottom-up, product-led growth model that had gotten Airtable to $11 billion was not sufficient, by itself, to win the largest deals with the largest companies.
So they built the top-down motion simultaneously. And the middle. And the edges. In trying to do too many things at once — Liu's own words, later — they'd grown the organization faster than the organization could absorb.
December 8, 2022. Airtable laid off 254 employees. Twenty percent of the company. In an email to staff, Liu acknowledged the failure directly.
"In trying to do too many things at once, we have grown our organization at a breakneck pace over the past few years."
He didn't dress it up. The company had moved fast in multiple directions simultaneously. The market had shifted. Economic uncertainty had arrived. The era of "growth at all costs" was ending for the entire technology sector — Airtable was not alone in this, Meta and Twitter and dozens of others were cutting in the same weeks — but the context didn't cushion the impact. 254 people were out.
The same day, the chief revenue officer, chief people officer, and chief product officer all departed. Three C-suite executives in a single morning. The leadership team was being restructured around a sharper thesis.
What that thesis was: enterprise. Fewer, larger customers. Not the scrappy marketing team building a content calendar for $10/month. The company that wanted to replace ServiceNow or Salesforce for a specific workflow category, and would pay a million dollars a year for the privilege.
"We want to consistently get customers with million-dollar-plus spend rates," Liu told Forbes, "versus supporting lots of little ten-thousand-dollar customers from a sales touch standpoint."
Nine months later, September 2023, the second cut. 237 more employees. 27% of what remained. The company that had employed nearly 1,300 people at its peak was becoming something leaner, more focused, more deliberately pointed at the enterprise.
"The market has tipped towards favoring efficient growth over growth at all costs," Liu said. "We are realigning to go after bigger use cases, and therefore bigger deals."
There's a particular difficulty in the position Airtable found itself in after the $11 billion raise. The money had come with expectations — not just from investors, but from the organization itself. The size of the round implied a size of ambition and a size of company to match. To take $735 million and then reduce headcount by 40% in the following year is to publicly revise the thesis. To say: we were wrong about the scale of effort required, wrong about the pace, wrong about how many fronts we could hold simultaneously.
Liu said it plainly. He didn't hide behind euphemism or restructuring language. The frankness was either admirable or the minimum expected from a leader whose company had just let go of hundreds of people. Probably both, in different measures.
What didn't change: the product. The 23-year-old's core insight — that most spreadsheet users were accidentally building databases, and deserved a tool designed for what they were actually doing — remained as structurally sound as it had been in 2013. The valuation had moved. The headcount had moved. The strategic direction had clarified, painfully.
The insight had not moved.
By 2025, Airtable was reporting approximately $478 million in annual revenue. The customer base had reached 500,000 organizations. The product had expanded well beyond its grid origins — automations, interface builders, AI-native features, integrations with hundreds of enterprise tools. The company had survived the correction and was growing again, more carefully, with more discipline, pointed at the market segment that could sustain a business of its scale.
Howie Liu was in his mid-thirties. Andrew Ofstad was still there, the cofounder who'd seen everyone's Airtable tab at WeWork and felt the world shift. The company they'd built — slowly, without pivoting, without chasing a niche, without compromising on the original architecture of the thing — was imperfect, and real, and used.
There's a version of this story that ends neatly, with the founder's insight confirmed, the product successful, the valuation justified, the journey heroic. That version is easier to tell.
The actual story includes watching a coworker type status updates by hand into a spreadsheet and thinking there has to be a better way. It includes three years of building something almost nobody outside the team could see the full shape of. It includes an iPhone app that ate eight months of work before anyone realized the product would spread through teams, not individuals. It includes watching the no-code movement you were building before it had a name finally catch up to you. It includes $11 billion on paper, and then a morning in December when 254 people got a calendar hold and a departure package.
The insight at the center of it was right. The observation holds: most people who open a spreadsheet are trying to build a database. They are using the wrong tool because it was the only accessible tool. Airtable made a better tool.
What the story of Airtable also says is that being right about the insight is necessary but not sufficient. You still have to build the right organization around the right product at the right scale. You still have to hire correctly when the money arrives and cut carefully when the thesis tightens. You still have to decide, twice in nine months, that you built too fast and must now cut what you built.
The spreadsheet-as-broken-database observation is still, in 2026, the best one-sentence description of why Airtable exists.
Howie Liu was 22 when he made it. He didn't know yet how long the road between the observation and the company would be.
Nobody ever does.
| Year | Event |
|---|---|
| 2009 | Howie Liu graduates Duke University (mechanical engineering + public policy) |
| 2010 | Founds Etacts (contact CRM); raises $700K including from Ashton Kutcher at YC Demo Day |
| 2010 | Salesforce acquires Etacts; Liu joins as product manager, leads social CRM |
| 2012 | Liu leaves Salesforce; Airtable founded with Andrew Ofstad and Emmett Nicholas |
| 2012-2014 | Two years in stealth; ~100 private alpha users; non-functional prototype → data-persistent product |
| 2014 | Invite-only beta launch on Hacker News; iOS app released |
| 2015 | Public launch (March); $3M seed; $7.6M Series A (CRV) |
| 2016 | Calendar view, kanban boards, gallery view; 2M+ bases created |
| 2018 | Blocks platform launched; $52M Series B; $100M Series C (Benchmark, Thrive) → unicorn ($1.1B) |
| 2020 | $185M Series D ($2.5B valuation); pandemic accelerates remote-work adoption |
| 2021 | $270M Series E; $735M Series F → $11B valuation; 300,000 organizations, 80% of Fortune 100 |
| Dec 2022 | 254 employees laid off (20%); CRO, CPO, CPeople all depart; pivot announced to enterprise |
| Sep 2023 | 237 more employees laid off (27%); full strategic realignment to million-dollar enterprise customers |
| 2025 | ~$478M ARR; 500,000+ organizations; AI-native features in development |
Howie Liu — CEO & Co-Founder. Duke engineering/policy. Built Etacts at 21. Sold to Salesforce. Left at 23 with the 90%-no-formulas insight and a decade of building ahead of him.
Andrew Ofstad — Co-Founder. Duke. Accenture R&D lab → Google Maps. Design and frontend lead. The one who stood in WeWork and realized the thing was real.
Emmett Nicholas — Co-Founder. Duke network. Palantir engineer. Backend architecture lead. Built the relational infrastructure that made flexibility feel effortless.
"Almost 90% of spreadsheets don't have formulas. Most people who use spreadsheets are building databases. They just don't have any other tool."
— Howie Liu, founding thesis for Airtable
