i gave a really fun talk last night on the general topic of saas, and the implications of 'the cloud' on the model. the deck is below, but the highlights (for what they are worth) are:
1. the saas model is the way to go (people have been saying this for 15 years, but connectivity and bandwidth costs make it real this time, not just another hype bubble) - i am a believer -
2. saas has fundamental advantages over traditional software - the obvious ones are distribution costs, but the less obvious ones are more flexible development and a greater ability to be focused towards 'agile' practices.
3. 'cloud computing' (which i will call haas) makes developing saas significantly more efficient and less expensive... developers can just do s and don't have to worry about procurement, modeling, etc for the h. this means that 'the cloud' will accelerate saas.
4. the proliferation of saas on haas, and even saas which consumes saas (via apis) makes it incredibly hard for end consumers to know what they are actually consuming - who has their data, where the points of failure are - etc. this is the polar opposite of the mission impossible 'clean room' in langley
5. this should open up huge opportunities for a new breed of web service security and privacy companies - specifically in the audit and risk management world... if these types of companies and services don't evolve the longterm potential for saas will be ultimately stunted.