Founding Family Values

Thoughts after the first four classes of Stanford’s CS183C
(ideas which surprised or resonated with me)

Validate early — any way you can:

make sure you’re building something people love

Paul Graham’s discussion of the “Manual” approach to “do[ing] things that don’t scale”—the “Wizard of Oz” approach in HCI lingo — reminded me of one of Tom Gruber’s excellent stories from the early days at Siri, pre-acquisition, back when it was a research project becoming a start-up. Early Siri was intended to take requests like “book me a table for two tonight at a romantic Italian restaurant in Palo Alto.” Before actually honing the underlying technology, Tom and team claimed it existed, and — in a hilarious inversion of the Turing test or ELIZA bot — they manually performed the Yelp lookups and OpenTable reservations as fast as they could after IMing “Processing your request…” to unsuspecting interlocutors. By the time Tom came into Apple (where he became my boss), Siri’s destiny as a household name — and perennial SNL topic — was basically secured. It was hard to imagine the early days, when Siri’s success was — as John Lilly described young Firefox or candidate Obama — so fragile and contingent. And it was fascinating to learn about how only through such acts of unsustainable efforts can anything lasting be built. Sam Altman’s story about the AirBNB founders was a testament to that.

My first attempt at a start-up failed. Like the Siri founders, I’d had what looked like a technological breakthrough in my research on computers and language that could help solve a real-world problem. But I imagined all the utility of the mature product before really validating that a market existed. As I learned about the space with my co-founders, it became evident that there really wasn’t a market for our software.

With Alation, we talked to would-be customers before we even had a product, and found many companies who had spent vast resources trying to internally solve the problem we were targeting (albeit not the exact same way). They were thrilled at the prospect of someone dedicated solely to solving that problem, since they’d looked around and found no solution they could buy. The validation (and ultimately revenue) from those early customers sustained us and helped us grow over the last three years.

Diversity and Unity

I founded my failed start-up with some of my best friends — extremely dedicated and talented people who happened to be some of the smartest I knew. The catch: selection bias. Geography is destiny and I’d met each of my then co-founders because we lived together (two in Stanford dorms, one on a ship). While we had complementary skills in engineering, linguistics, design, and management, not one of us was over 25. Age doesn’t matter, but experience can. None of us had run a company, executed an enterprise sale, or built a widely-used website. In Silicon Valley, we love to point to the outliers who irrevocably changed the world as college drop-outs with no prior experience, but statistically speaking, that’s not the most likely outcome. (Happy ending: we’re all still good friends, and they’re kicking ass at Facebook, Uber, and MIT, respectively).

By contrast, I didn’t know my Alation co-founders at all before Alation. But we had each independently developed a contrarian view of the key problem in our space, and we had each independently envisioned a similar approach to solving it. We shared that vision, but differed totally in our backgrounds. Superficially, we came from three different countries and were born in three different decades. More importantly, we had deep and deeply complementary skills in business and sales, Google-scale engineering, and product design. It felt like “Mission: Impossible.” That breadth of skills is hard to find in a single college cohort, even at Stanford.

Sam Altman described the ideal founding team perfectly: diversity of perspectives is really good. Diversity of vision is a problem.

A pleasant surprise

I came into this course with pretty high hopes. So far, those have been exceeded. I did worry that maybe the content would be a string of generalizations or platitudes. But each instructor and speaker has provided such great examples and strong perspectives, and displayed such wit, compassion, and humor. I’ve left each class feeling energized, inspired, and educated. Looking forward to learning more, especially as we discuss OS2 — of particular relevance to Alation at our present stage.