Analogous estimation is a form of top-down estimate that makes use of the actual cost of previously completed projects to predict the cost of the current project. If the project being used as the analog of the project being estimated is a close match, the estimates could be quite accurate; on the other hand if the purported analogy is more apparent than real, the estimate might not be very accurate at all.
This point is crucial. It frequently happens with software development projects that many analogous previous projects can be found and examined, often sharing an apparently similar design of many of the key code modules. At first it might appear that if the difficulties of the projects are similar, then a 30-per-cent greater size in total program code of the new project should lead to a 30-per-cent greater cost than the analog being taken as the benchmark for comparison. However, if programmers’ time-saving code productivity tools have improved in the interim, the new project could cost less even though it uses more code. Alternatively, if the previous project was using some of these newer productivity tools but those tools are not yet part of the planning of the current project, a 30-per-cent increase in coding will actually cost the new project more than an incompletely-informed analogous estimate would predict.