Smart startups have to think about how to take the friction out of their business model. Primarily that means looking at your value stack, and figuring out a way to take human beings out of your process that slow things down. The good news is that more and more customers hate dealing with people, as they are analog, slow, and imperfect. Some other more mundane examples: Dell product ordering, Walmart procurement, Intel and Cisco fufillment, online travel services, grocery self-service lines, and on and on.He doesn't necessarily suggest that taking the human out of the picture is the sole opportunity, but rather simply taking friction out of the system, creating value, and sharing it with the customer.
Effectively, Steve is championing the idea that huge homeruns are created by (a) having very scalable business models that (b) don't require much (if anything) to complete an advantageous sale and/or (c) enable mass customization. Products or services that require hefty sales forces and lengthy sales lead times are simply an enormous impediment to certain explosive growth. In an established market space, such as the one that Google broke in to, market scale was created by automating sales and enabling SME's to buy the same products that were only really available to large enterprises.
Skype has grown because it simplified VoIP across firewalls and enable NAT traversal. Intuit grew because it made software easy to use for the SME. Inphonic grew because it simplified infrastructure for wireless carriers and MVNO's. AOL exploded because it made the Internet easy. Microsoft took off because they made (after Apple's idea) the personal computer's GUI straightforward. Bloomberg boomed by making information instantly accessible. And on. And on.