What I meant by that statement is that as an application owner in CloudFoundry you get to decide how many resources are brought to bear to run your app. In particular, you get to decide how many instances are active, and presuming an instance-hour cost model this gives you control over the application cost. This means that you get to decide where you want to be on the cost/performance curve, e.g. whether you want to deploy more instances to improve performance (assuming this actually scales performance) or whether you'd rather save at the expense of performance. Also, if there is unwanted traffic hammering your application, you can decide not to scale-up and thus not to spend extra money on servicing these unwanted requests. This is in contrast to PaaS services where you pay essentially per-request and/or where you cannot easily control the scaling. Or where cost control comes in the form of a quota with the result that your app is "off the air" once the quota is reached.