As suggested by Tables Table 31-1 and Table 31-2, the bottom line on RDLs is that the specific savings you can expect depends both on the specific language you're using now and on the specific RDL you switch to. It also depends on whether the kind of program you need to build is the kind the RDL is good at building. If you're currently using a 3GL, you can probably expect to cut construction effort by about 75 percent when you switch to an RDL. That amount will continue to improve as new and improved languages become available—but it probably won't improve as quickly as tool vendors will want you to believe!

If you can't switch completely to an RDL, you might still be able to implement some of your project in an RDL. The 75 percent rule of thumb applies: you can expect to cut your design and construction effort by about 75 percent for the part of the code you implement in the RDL (Klepper and Bock 1995).

As project size increases, the savings you realize from switching to an RDL decreases. RDLs yield their savings by shortening the construction part of a project. On small projects, construction activities can make up as much as 80 percent of the project's total effort. On larger projects, however, detailed design, coding, and debugging shrink to something like 25 percent of the total effort, and the savings from streamlining those activities will shrink accordingly (McConnell 1993).