I know everything is moving to SaaS for very good reasons (both for the developers and for the customers), but one thing the old packaged software business had

I know everything is moving to SaaS for very good reasons (both for the developers and for the customers), but one thing the old packaged software business had going for it was that costs were totally decoupled from pricing: your largest cost tended to be development, which meant that you could (in theory) just sell the same thing to 10x the number of people without substatially increasing your costs. Ironically, at least in the enterprise space our experience is that companies will pay much more for installed software than for SaaS, which seems totally counter-intuitive. When you're selling them installed software, they look at it as an investment, evaluate ROI, compare against the amount required to do an internal build, etc. When you're selling them SaaS, they look at it as a cost center to be minimized, and their interest is primarily in driving your costs through the floor. To put it another way, with installed software they realize they're paying your development costs, and that even if 100 other people are as well they know that it's still cheaper to buy from you than to build themselves. With SaaS, they act like they're paying you for hosting and support and totally ignore the development cost aspect of things, so their acceptable price is much lower. Irrational? Sure. But when my company investigated doing a hosting offering of our enterprise insurance products, we basically saw that it would probably end up just lowering our margins significantly, as well as probably lowering a potential future market valuation. Any sales we lose from not having an official hosted offering (we have partners that will host our software for them, but they're still paying us the full license fee) are ones we're willing to live with. My understanding is that public markets have tended to value companies with that exponential profit growth ability at higher valuations than companies with more linear, fixed margins. There aren't all that many SaaS business public at this point, so it'll be interesting to see how the market values them once they are; my impression is that equivalent installed-software and SaaS businesses (same revenue stream and profit, say) will have wildly different valuations, fair or not, due to the theory that the installed-software vendor can much more easily turn the crank to increase revenue without increasing costs.