tl;dr: Don't worry about your employability, but your startup has a bigger chance of success if you do the big company <i>first</i> and then the startup.<p>As long as you challenge yourself during the startup, it's normally not a problem. Typically, having "Founder" on your resume attracts a lot of attention, and then the purpose of the interviews becomes to determine whether you actually have the skills that a founder should or whether you just sat back and waited for the phone to ring the whole time.<p>A word of warning: when I got out of college, I read all the articles that said "Go blue-chip", figured they didn't apply to me because I only ever wanted to do startups, and so I worked at a startup for a couple years and then founded one. And while I don't exactly regret my choice - I needed to do it, based on my personality and my experiences in life up to that point - I think it made things harder on me and set me back a couple years.<p>You get several things "for free" when you take a job at a big, successful company. You'll pick up development practices like version control, continuous integration, unit testing, etc. because they're just the way things are done. You'll probably learn several technologies, though those technologies are often more proprietary than you'd get if you struck out on your own. You'll learn how large software systems are put together. You'll learn the sorts of problems that companies with money face, which is absolutely essential if you ever want to sell to said companies. And you'll hopefully build up a network of trusted coworkers that you can tap for cofounders later.<p>All of that stuff, I had to teach myself when I was working at somebody else's startup and later founding my own. I remember scouring the web for any public information I could find about how Google put their infrastructure together, and then when I actually got to Google, I found that the publicly available information is only the tip of the iceberg. A/B testing wasn't widely publicized at the time; nowadays, most people in the startup community know about it (largely thanks to patio11), but when I was hunting around, the only indication I had of how this was done was Max Levchin's 2-word aside in a PayPal presentation: "Measure everything". I remember obsessing about things like scalability and internationalization that really aren't that important for a startup, but I vaguely knew that successful companies did them, so I had to know how.<p>Most of that becomes a "well, duh" once you've spent a couple weeks working with it in practice. At good companies, there'll be code review, so you can just have a senior engineer tell you how it's done. And then once you do strike out on your own, you can say "Well, I know how to do this, but it's a gigantic timesink and not really that important to my current situation, so I'm going to leave it until later," and focus on the parts that actually are important, like getting product/market fit.