I haven't been able to find much about the technical lessons learned. I found lot of high level non-technical and marketing advise.<p>It would really be interesting to get feedback as to what technical things did some of the founders learn.
I was a hacker edged into the biz/sales role (we had 2 other, stronger tech guys on the team). So this is my perspective from watching (and building pieces of) the tech over 3ish years.<p>Code quality matters. We did enterprise sales across several products, and it got to the point where making minor tweaks to the older products to close a sale was a complete nightmare.<p>You probably don't need to rewrite it. In most cases I've seen, re-writing (e.g. switching technologies) is some sort of pathological denial about delaying the admission that nobody wants the product.<p>The importance of scalability depends on your channel. We spent a lot of time building in scalability, which was important for an advertising company (we couldn't close sales w/out scale guarantees), but might not be important for a word-of-mouth app.<p>Build an admin dashboard. A good one. You're losing countless weeks if the CTO is a bottleneck for deploys and minor database tweaks.<p>I wish we'd used a modern web framework. I didn't write any of the server side code, but our development felt particularly sluggish there.<p>Check for plugins and tools before writing it. Some of the stuff we wrote from scratch was 2x commenting systems, 2x moderation systems, 3x analytics & charting systems, and a widget sharing wrapper. Writing some of those in-house was correct, but we did it somewhat compulsively.<p>Think long-term. We wrote some widgets in Flash, then switched to Flex because development was slow. But for our ultimate vision, we needed <10kb widget sizes, so we had to switch back to Flash a year later. Since the core of our product was in that widget's UI, it was a non-trivial rewrite.<p>Write a stupid version of the UI first. A major UI overhaul delayed the rest of a key deploy for ~3 months. Starting with a quick v1 UI which copied twitter would have allowed the UI/UX guys to do things properly without the server team screaming at them every day.<p>Shipping makes people happy. If you have an API, get people writing (& launching) little apps which use it. Have them own those apps personally and keep them on their personal githubs/portfolios. If you are playing in an industry, encourage people to make 1-2 day apps, separate from your product but wthin the same space. You'll stumble across some gold.