100% agreed. I've participated with a startup/think tank recently (<a href="http://webecologyproject.org" rel="nofollow">http://webecologyproject.org</a>) and while we could have had offices- we didn't. It worked so much better for the most part.<p>Our downside to this was twofold however:<p>1) We choose our team poorly & it was too big. That isn't to say that these aren't all simply wonderful people, but for making a highly efficient machine- it was a hodge podge of people with various levels of interest, skills, experience and commitment. We crept up to 10-12 people quickly and since it wasn't a formalized business structure at first this seemed ok. We were however all in the same geographic location (Cambridge).<p>2) We had too many meetings, which people stopped attending and become ineffective. These became a time sink. People would defer conversation to the meetings instead of our internal email list (which for some conversations is the right thing to do), but then these became lengthy 4-5 hour meetings that stopped getting things done. Our coding slowed. Decisions stopped being made, interest was lost slowly.<p>These are points very specific to our group and I totally agree on holding off on office space until you MUST have it. Keeping costs down and excellent workflow up initially are a must. Just don't try it with 10 people and don't kill people with mandatory meetings. Use IRC or some other tool religiously.<p>If anything, look into a coworking space like Betahouse (<a href="http://betahouse.org" rel="nofollow">http://betahouse.org</a>) or New Work City.