**Parametric modeling**- Uses a parameter, such as cost per metric ton, to predict project costs and relies on knowledge of mathematical relationships between two or more characteristics of a project. For example, accurate cost estimates can be obtained in house construction through historical knowledge of the statistical relationship of dollars per square foot/meter. Similarly, one might collect information on dollars per line of code for software development projects or on dollars per lane mile/kilometer in highway construction. Parametric estimating uses heuristics, rules of thumb to estimate project costs. Parametric estimating is usually faster and easier to perform than bottom-up methods.

**Note:** A real world example of a parametric model is called the learning curve. The learning curve mathematically models the intuitive notion that the more times we do something the faster we will be able to perform the task. Specifically, learning curve theory says that each time we double the number of times we have done something the time it takes to perform the task will decrease in a regular pattern.

- Bottom-up estimating - Starts from zero and adds the expenses from bottom-up. Bottom-up cost estimating involves doing detailed cost estimates for project work packages and then aggregating the results up through the cost accounts and higher levels of the WBS in order to arrive at an overall project cost estimate.
- Top-down estimating - Uses a similar project as a cost baseline and factors in current project conditions to predict costs.
- Rough Order of Magnitude (ROM) estimates are approximations without detailed data, often done early in a project when you need a “ballpark guesstimate.” Used in top-down estimates. The range of variance for the estimate can be –50 percent to +100 percent. ROM estimating happens during the initiation phase of a project.

**Budget estimate**

This estimate is somewhat broad and is used early in the planning processes and also in top-down estimates. The range of variance for the estimate can be –10 percent to +25 percent. Budget estimates are made during early planning processes. Budget estimates are based on slightly better data and are often used to establish initial funding and to gain project approval. The range of accuracy is -10 to +25%

**Definitive estimates**

This estimate type is one of the most accurate. It is used late in the planning processes and is associated with bottom-up estimating. The range of variance for the estimate can be –5 percent to +10 percent. Definitive estimating happens during late planning process. Definitive estimates are prepared from well-defined, detailed data. A bottom-up estimate would be an example of a definitive estimate, which is the most accurate (with a range of -5 to +10%).