Estimating the Data Warehouse Project: Wild Guess or Exact Science?
Unfortunately, estimating the duration and effort of a data warehousing project is not a precise science by any stretch of the imagination. Too many times have I heard the words: "six-month turnkey " or "out-of-the-box" solution. These solutions are never six months and are never out of the box.
To properly estimate a data warehousing effort takes experience, a deep understanding of what is being built and pure gut feel. There are numerous factors that impact the schedule on a data warehouse project. In this article, I'll discuss some of those factors and give you some guidelines for estimating the data warehousing effort.
The most important criteria in estimating the duration of a data warehouse effort is the number of sources being integrated into the data warehouse. As a general rule, I would not integrate more than five sources of fact data into a data warehouse during one project cycle.
My experience has shown me that you can have five developers manage the process for five sources concurrently, but beyond that, it becomes extremely unwieldy to coordinate and manage the integration of the data. It takes about three months to integrate each new data source from design to implementation per ETL developer once the data model is already in place.
Note: it does not take 1.5 months if you have two ETL developers for one source.
Please log in or sign up below to read the rest of the article.




