Its great that the customer believed in you and your product so much that they basically funded you by buying initial versions that were "bad". What was the product? Were there few or no competitors in the market? I assume the company which purchased your initial versions would have just gone with your competitor otherwise.
The product was BinNavi (http://www.zynamics.com/binnavi.html) which sold for (I believe) roughly 3000 Euro per license despite being really terrible for the first two years of its existence.
The only competition was IDA Pro (http://www.hex-rays.com/idapro/), the standard tool for binary code security researchers, but BinNavi aspired to be much more than IDA Pro and our customer was aware of our vision and bought many licenses to support us.
The company charged hundreds and thousands of dollars for software licenses from day 1 even though the first versions of the products were pretty bad. In fact, we even had customers who bought these early versions not because they wanted us to succeed. I remember at least one customer who bought licenses for many tens of thousands of Euro telling us 'we won't use your software yet because it is not good enough but we believe that you will deliver great software one day and by buying early version we will keep you alive'.
So yeah, charging right from the start worked out tremendously for us. Without it we would not have survived.