Databases Reference
In-Depth Information
Step 1: decide what to analyze, measure, or forecast
Business decision: What questions do we need answered? What is being measured that can provide
answers to critical questions we have of the data?
We thought it a little funny that there is a person sitting down for Step 1 in Figure 3-6, as if she is
waiting for her report. In many cases the business users are accustomed to getting static reports and
perhaps do not recognize their value as a stakeholder or their role in the iterative process of deter-
mining what to measure. Additionally, often things are being measured, but the business users don't
know how or where to get the data—and if they get it, it is a cumbersome manual process. Addition-
ally, they might or might not realize that the result of their efforts will produce interactive reports
rather than static reports, in which case their role is to become more analytically minded by asking
questions of the data—to then further the design process for data-driven reports that directly affect
decision making.
There are many possible scenarios that drive reports. For example, your company might decide
that it wants to answer questions by using data it has collected from transactional systems, or the
company wants to improve its forecasting process by collecting the right information by having more
reporting flexibility or by giving more people access to information. For example, a common request
is to combine transactional data with marketing campaign data. Another is to pull in data from the
cloud that is related to demographics or weather. The required steps for these BI initiatives can go
wrong in several places. You need to get to what your customer wants, and you want to be effective
so that you build the solution that the organization needs.
Discovering and exposing data and determining what to do with the data might be the most
important and often the most difficult steps. It's important because you do not want to spend
resources collecting data that is not useful. Savvy developers and solution designers must work with
users to determine their data requirements in an efficient, iterative manner.
Be patient if they do not know what or how they want to measure. Or, they do not know if there is
trusted data available. Deliver prototypes and give them an opportunity to explore data and visual-
izations. It can take time, but you will gain their trust and they will begin to trust the data.
In our corporate BI scenario, the easier a designer or modeler makes it for users to quickly under-
stand the results of a BI solution, the closer a designer can come to delivering a useful solution. For
this reason, one of the multidimensional tools, the SSDT browser, gives developers the ability to see
the results of data collection and queries, and it is nearly identical to the viewer in Excel. Additionally,
you can view your tabular data results in the Excel client in SSDT.
Finally, because you can import your tabular project from PowerPivot in Excel 2013 into an SSDT
project, consider using PowerPivot or Excel to prototype and to give users an opportunity to explore
the data before formalizing the data processing or refreshing.
Each of these tools provides great support for prototyping and scope checking.
Search WWH ::
Custom Search