My first Publication Agile-Data-Warehouse-Design-eBook | Page 136

Modeling Business Processes
115
Using the Matrix to Find Missing Events
When PRODUCT RETURNS has been added to the matrix, you can check for a
missing event in the sequence by asking the obvious question: “Does anything
happen after PRODUCT SHIPMENTS but before PRODUCT RETURNS?”
Alternatively, you can get stakeholders thinking about what might belong in the
gap, if there is one, by asking:
Check for missing
events by looking
for gaps on the
matrix
Does anything difficult, costly, valuable, or time-consuming
happen between shipments and returns?
This might prompt stakeholders to think of CUSTOMER COMPLAINTS. If they
agree that this event represents the start of a new process you would add it to the
matrix as in Figure 4-14, which now shows PRODUCT RETURNS indented as a
milestone in that process. Notice that the exceptional step brackets have been
removed from PRODUCT RETURNS. Not every complaint leads to a return; for
example, a complaint might be that a product hasn’t been delivered yet (another
event: “Carrier delivers Product”, to add to the matrix), but a high enough percent-
age of complaints do result in returns, enough for stakeholders to view return date
as a standard milestone of this new customer support process.
Look for large time
gaps or value
changes. They often
represent the start
of a new process
Figure 4-14
CARRIER
DELIVERIES,
CUSTOMER
COMPLAINTS and
SALES TARGETS
added to the matrix
Trying to find the correct position for an event within a process sequence can often
help to expose additional events that represent the end of one process and the start
of another. In our example, deliveries are the final milestones for most orders.
Complaints and returns, on the other hand, are thankfully not part of many orders.
The indentation in Figure 4-14 shows how CARRIER DELIVERIES completes the
order fulfillment process and CUSTOMER COMPLAINTS begins a new customer
support process. Documenting the first and last events of a process is particularly
important. They represent cause and effect, origin and outcome and are the most
Model the first and
last events in a
process. They are
the basis for almost
all process
performance
measurement