My first Publication Agile-Data-Warehouse-Design-eBook | Page 144

Modeling Business Processes
123
Figure 4-20
Adding details to the
shipment event
Following the where details, you ask for the how manys. These quantitative details
do not feature on the event matrix, just like the when details, and are the main
reason for modeling the event in table form; the matrix shows how events are
described using (conformed) dimensions. The how many examples show how
events can be measured. Ask how many? to
In Figure 4-21, two new quantity details: SHIPPED QUANTITY and SHIP-MENT
COST have been added, along with the ORDER ID why detail. The quantity
examples are new, supplied by stakeholders, but the order ids are copied from the
previously modeled CUSTOMER ORDERS event because ORDER ID was identi-
fied, on the matrix, as a conformed degenerate dimension linking the events. With
that existing why filled in, you ask for additional whys, remembering to ask why
quantities vary. If you know that an event, such as shipment, is a process milestone
you should ask why similar details vary (or do not vary) within the process; for
example, you might ask: Add any existing
discover the event
measures not
modeled on the
matrix
why dimensions
from the matrix
and ask additional
why questions to
explain story
variations in the
measures
Why can SHIPPED QUANTITY differ from ORDER QUANTITY?
and get the answer:
Several partial order shipments can be made when product
stock is too low to completely fulfil an order line.
This tells you there is a 1:M relationship between orders and shipment. It also tells
you that you haven’t yet found a combination of details that would make a ship-
ment event unique. You record this by adding a new repeat story to the table, as in
Figure 4-21, which demonstrates there can be multiple identical shipment events
for the same order line item by duplicating the granular details (ORDER ID,
PRODUCT) of an original order (ORD5466).
Why answers can
represent the need
for additional
examples as well as
new why details