Startup, in detail
Startup is Jerry Kaplan's account of GO Corporation, the pen-computing company he co-founded in 1987 and ran until it collapsed in 1994. GO raised over $75 million, attracted engineers from Xerox PARC and Apple, and was convinced it had built the future of computing — a tablet-based pen interface that would replace the keyboard-and-mouse paradigm. Microsoft and IBM both tried to kill it. AT&T eventually acquired it, then abandoned it. The book tells that story with unusual candor.
What makes Startup distinctive among tech failure memoirs is Kaplan's willingness to describe the social machinery of Silicon Valley in uncomfortable detail. The fundraising meetings, the board politics, the moment when a venture capitalist signals approval not with words but with a change in body language — Kaplan captures these dynamics with a novelist's eye for texture. The reader sees how deals actually get made, how loyalty shifts when the cap table changes, and how a company can be technically ahead of its time and still fail for reasons that have little to do with technology.
The pen computing thesis wasn't wrong, exactly — it was roughly a decade early. The iPad arrived in 2010 and validated most of GO's core ideas. Kaplan is good at analyzing this particular kind of failure: not stupidity or fraud, but a correct hypothesis deployed in the wrong window. He argues that timing is a strategic variable in the same way that product quality or distribution is, and that Silicon Valley's mythology of the visionary founder systematically underweights it.
The book reads fast and the characters are vivid. Bill Gates and Andy Grove appear as they actually behave in competitive situations — not as villains, but as formidable players protecting their own franchises. The ending, when Kaplan describes the final dissolution of GO and the dispersal of its team, is genuinely affecting. For anyone trying to understand how technology startups live and die, Startup remains a clear-eyed primary source.
The big ideas
- 1.
Timing is as important as technology. GO had the right product vision but arrived before the hardware and networks could support it commercially.
- 2.
Venture fundraising is a performance with its own social codes. Reading the room — literally — matters as much as the deck you've prepared.
- 3.
Large incumbents (Microsoft, IBM) don't need to build a better product to kill a startup. They need only to create enough uncertainty to dry up investment and customer confidence.