My first Publication Agile-Data-Warehouse-Design-eBook | Page 146

Modeling Business Processes
125
Completing the Event
Copy the how details SHIP MODE and SHIPMENT NUMBER from the matrix to
the event table and ask for examples—as neither has been modeled in table form
before. The SHIPMENT NUMBER examples should confirm that you have finally
found a detail that can differentiate two partial order shipments of the same prod-
uct to the same customer on the same day. With this final piece of the puzzle you
can define the granularity and type of PRODUCT SHIPMENTS. The granularity is
a combination of SHIPMENT NUMBER, ORDER ID, and PRODUCT, and is
recorded by marking these details as GD (Granular Details/Dimensions). From a
business perspective this granularity can be described as “Shipment note line
items”. This granularity makes the story type DE (discrete event). Figure 4-22
shows this information added to the final spreadsheet version of the event table.
Complete the event
table by recording
event granularity
and type
When you finish modeling an event table don’t forget to model dimension tables
for any details that you have marked as [?]. You still need to define some dimen-
sional attributes for these details, before ending the modelstorm.
Figure 4-22
Completed
PRODUCT
SHIPMENTS event
You might be tempted to start a modelstorming session by using a matrix to
model and rapidly prioritize multiple events. While this can work well with stake-
holders who are already familiar with the process, it can be too abstract for some
brand new modelstormers. Remember many BI stakeholders would prefer to
define reports rather than data models. Starting with an event table and example
data (even if it is not for the most important event) that looks like a report can help
stakeholders get the matrix and appreciate its value.