[under constru

I've been kicking around a way to codify project estimates based on common or typical CRUD (CrudScreen) artifacts. Not every part of a project will fit into CRUD-based estimates, but often large portions of it will.

An entity (table) is the starting point. A "multiplier" may be needed for different languages and tool stacks. For example, using Oracle or DB2 will usually take more setup work than using MySql or MS-Sql-Server. The numbers below are merely a starting baseline or template to be adjusted based on experience and shop practices.

If a characteristic says +10%, it means the labor estimate is 10% more. Thus, if the baseline entity is 20 hours, then +10% adds 2 hours. The "add ons" are not cumulative unless stated otherwise.

'''Entity'''

Starting point is 5 hours of labor per entity, plus a half hour per basic column, 1 hour for basic look-up (pick-list), or 2 hours if column needs complex look-up, such as QueryByExample.

+60% if entity is write-able via your project (versus using some other organization's or department's data).

+30% if QueryByExample will be used (versus simple lists).

+20% for each report format, +35% if the report has more than 20 fields/columns. +15% for each level of sub-totalling.

+30% for each additional "edit view", such as having different screens per user role.

+10% for each additional column in entity's primary key, if compound.

--top


------
CategoryProject