A startup’s role is to discover what customers want/need, then tweak and tune their product and sales models to create an effecient process to signup more customers (thus scale).

Roman Stanek’s Nov 1 post on TechCrunch “Avoiding the Cargo Cult and Getting the Trans-Atlantic Startup Model Right” http://www.techcrunch.com/2009/11/01/avoiding-the-cargo-cult-and-getting-the-trans-atlantic-startup-model-right/ contained this paragraph:

“I am a huge fan of the Customer Development Model by Steve Blank, but it assumes that the company can continue spending money on engineering and market/product fit tests until the target market is actually validated. And as much as I like the startup ecosystem here it seems to me that the people cost of software development forces startups to launch half-baked products. Very few companies can make the “Four Steps to the Epiphany” work financially – this is one reason having engineering located in a cheaper country from day one is a major plus.”

Which seems to offer an understanding of “customer development” that’s quite different the Four Steps to the Epiphany approach. The Customer Development Model is designed to minimize the risk of launching a product for a market that does not exist. It encourages a startup to invest in customer discovery and validation in parallel with product development prior to product launch.

Lowering/increasing the cost of engineering personnel doesn’t affect this model (although the trend over the last two decades has been to dramatically lower the cost of developing a software product: open source libraries, Moore’s Law, Internet distribution/SaaS are three big drivers).

As your diagram makes clear: there is no point in suspending your efforts to validate product/market fit in favor of “launching” unless you have evidence of demand and a repeatable scalable process for creating customers.