It was a really slow transition, that went something like this:

Phase 1: I did all the design, front-end development, and client management.

Phase 2: I did all the design and client stuff, but I started handing off front-end development to contractors.

Phase 3: Still did all the client stuff, but I focused on just doing the high level design, then worked with other designers and developers to finish the work.

Phase 4: Fully handed off all the client relations to someone else. I still creative directed and did meetings with clients, but I was much more hands off in terms of the actual pixel level design work.

Phase 5 (Today): I found an amazing CEO to actually run MetaLab day-to-day. I focus on high level strategy and chime in on design projects on a high level. I loved running the company, but I was trying to run multiple companies at once and it ultimately just broke down. I find that I'm still able to have a major impact on a lot of the work that goes out, while leaving the day-to-day running of the company to someone else. Loving it so far.

In terms of my advice, my friend Jason Fried from Basecamp wrote a great piece about giving advice. My perception of how I got here is probably totally inaccurate, and my advice probably doesn't apply to today's world.

That said: entrepreneurship is really just delegation. I just kept progressively delegating more, and more, which allowed me to grow my company, then eventually multiple companies. It took 9.5 years, so it's just a slow and steady process of delegating little bits here and there. One day, I woke up and I was like "holy crap I have 120 employees and four companies". Wat.

If you ultimately want to do design work every day, and stay heavily involved in the day-to-day pixels, then scaling will be challenging and you will probably want to stay smaller. If you relish the idea of focusing exclusively on the high level, then hop on the delegation train :-)