Software Size Estimation Technology Trend

SEER supports numerous logical and physical sizing methods including but certainly not limited to Source Lines of Code (SLOC) and Function Points. We have found alternate sizing metrics to be particularly useful early in a project when less information is available. We have also found that customers without an existing in-house sizing standard benefit from alternate metrics. We therefore not only support many alternative sizing metrics out of the box, but have added several features that relieve the estimator from knowing anything about a specific size.
Galorath distinguishes itself by the way its tools can and frequently are adapted to customers’ unique needs. As this section illustrates, sizing methods may include traditional metrics, custom ones, sizing by analogy, by past history or a mix of these methods.

Delivered size metrics include:
•Source Lines of Code
•Function Points (IFPUG, COSMIC, Mark II)
•Detailed Function Based Sizing (Screens, interfaces, reports, etc.)
•RICEF Objects (for ERP systems)
•Data Warehouse/Business Intelligence Components (ETL operations)
•COSMIC Function Points
•Use Case Points
•Story Points
•Number of requirements
•…and any other effort-driven sizing approach can be defined and used

Fingertip Access to Size Metric Choices in SEER-SEM

Function Based Sizing Galorath has consistently innovated in providing more usable sizing options. For example, our own Function Based Sizing metric has long been widely used to make software sizing easier. It doesn’t require detailed specifications nor does it require detailed training in function counting. If you have only general specifications, and a generally reasonable description of how the finished software will function, you can still produce a highly useful size estimate using FBS.
Function Based Sizing uses the main IFPUG function categories decomposed into simple functions such as input screens, printed reports, and menus. It tells you what to enter in plain, straightforward language, organizing the inputs in a coherent, logical format.

Function Based Sizing Makes Sizing Intuitive

Story Point Sizing Example

Estimate Based On SAP RICEF Components

SEER allows the user to mix sizing methodologies within the same project and even within a single WBS element —very useful when a project is comprised of several different technologies and development methodologies. You can also switch the metric used to size a project, should another later become more appropriate.

All sizes can be described as new or preexisting. Rework of existing code is measured to compute an effective size. The SEER model utilizes an effective size which normalizes programming languages and provides a single size measure for all projects.

Estimate Sized Using Existing SLOC and New FP Counts

An estimate prepared using multiple sizing techniques (Function Points, Lines of Code, Use Cases, Story Points, etc.) is normalized into effective effort units. This common measure provides the continuity required to perform multi-component comparisons and to obtain a common productivity measure.

Analogy sizing dispenses with cardinal measures entirely by asking users to intuitively select whichever analogy best describes a current project.

Analogy Based Sizing – Sizing Using In-House Nomenclature

Custom size metrics map whatever terms and quantities a user is familiar with to (within the model) lines of code and function points. This method mates user intuition with proven size metrics. Galorath regularly develops new custom size metrics at customer requests; in fact, the number of different entries in custom metrics has recently been greatly expanded due to one such request.

Defining a size metric is a straightforward process using the Size Metric Definition feature.

Creating Size Metrics

SEER-SEM allows you to build calibrated size metrics based on known task effort. The size metric feature allows a relationship to be automated between the effort hours to perform a task (such as writing a report) and an equivalent SLOC or Function Point value. The model includes a predictive algorithm using productivity and effort to generate the size equivalents.

Task Based Sizing Allows Estimation Using Standardized Task Times