Every company dreams of getting larger and having more customers (hopefully capturing more value out of each transaction). Unfortunately, uncontrolled growth can disrupt your finely tuned-processes, it can overwhelm your Customer Support infrastructure, and thus, creating a situation where you are growing but dropping in customer satisfaction and repeat customers (it happened to Dell a few years back). So, what is the right choice? Control your growth? or Grow and fix things latter?
If you just grow and expect to fix things latter you may be in for a surprise: bad habits will have set in, your reputation may be already tarnished, your competition will be runing ads making fun of your service, and your customer will be looking for the next new thing! Internally, because of your hurry, you didn't build things the right way, you lost good but overworked persons and lost employee goodwill. Fixing things will cost you a fortune. I believe myspace would have been the Facebook of today if they had manage their growth better and kept their ears to the market.
Controling your growth may be the better answer. It allows you to keep your existing customers happy, while making shiny new customers. The difficult question for me is: what is the optimal speed for growth? The way to gauge your optimal growth speed is to use your "ears" to the market to determine when you are sacrificing quality for growth and shift your focus accordingly (hint: it isn't easy).
This is where your Product Management and Customer Service organizations are invaluable in getting a readout on the market (if you don't know what your customers think about you, you are in trouble). You need to be bold and put new infrastructure in place to support your growth, but as soon as customer satisfaction drops, you need to refocus on getting things under control again. If you keep doing this as fast as your organization is capable of, you are getting the best of both worlds. Hopefully, you will be fast enough as to not give your competitors a window of opportunity.