Meh... all generalizations are flawed (including this one?) Seriously, I think <i>any</i> blanket assertions about whether a startup should work out of a co-working space or not, or whether they should raise money or not, or whether they should use Rails or Grails or Node.js, or whether they should "X" for any value of "X", are probably wrong as often as they are right.<p>Regarding co-working spaces... We have a co-working membership at the Underground @ Main in Durham, and mostly work out of there, Beyu Cafe, Barnes & Noble, Panera Bread, and/or our respective homes.<p>Our own door? Feh, if the bar were set that high, we wouldn't even exist and be doing this, since there's no money to rent office space right now. And any money we <i>could</i> spend on office space is almost certainly better spent on something else.<p>My point is, context matters. If we wind up raising money at some point, and/or get some serious customer revenue coming in, then yeah, sure, we'd start looking at office space. But right now the UG@Main works great for us. There's a nice place to work with fast 'net access, convenient access to the rest of downtown, AND we can reserve a conference room if we need it for a customer meeting, investor meeting, partner meeting, etc. TBH, that's the main reason we even have co-working here, so we'll have a place to meet people that's better than saying "Hey, let's meet at Starbucks".<p>My feeling: Get an office when you're ready for it, not sooner. :-)<p>Edit: to expand on this... I almost think getting a "real" office too sign could be a sign of something like "cargo cult thinking". I mean "real businesses have offices, so we need an office so we'll be a real business". Maybe it's closer to "take it 'till you make it", but office space tends to be pretty expensive, and my thinking leans towards avoiding that expense as long as possible. Heck, as far as that goes, maybe it's time to ask if businesses need office at all. Why not go fully distributed, 100% tele-work?