So, your boss told you to prepare interactive dashboards related to your business that could be easily shared with other
colleagues and all based on data in your SQL and OLAP databases. The weekend has gone too soon and now you have to think of
something while your boss is still drinking Monday coffee. Finally you remember what you've heard about 'Kyubit BI' and decide
to give it a try. You will learn that coffee time leaves enough space to build dashboards with this software.
Well, hold on tight...
Download
Download of about 10MB of complete setup program (not only setup starter) will be ready in seconds.
Well, you need to fill in the download form with your name and email address.
Setup
Setup requires some Windows components to be active, such as Internet Information Server and ASP.NET.
If you do not have those features enabled, it will take you an extra minute to configure from Windows features,
but don't lose your hope, the rest of the setup will be done in a minute. After setup is done, the app will be
automatically open in your browser, web based and ready for work.
Add Data Sources
Add Data Sources that you will use to connect to your data. Fill in a simple form with connection details to make
permanent data source definition for your SQL, OLAP, and ODBC data sources. Optionally share data sources with
other users. You can immediately tell your colleagues to start using data source definitions for their own analyses and dashboards.
OLAP analysis
Create OLAP analysis by dragging and dropping dimensions and measures to analysis grid view, categories, and series axis.
Select any number of measures and filters to narrow analysis insight. Search or browse for required dimension members.
Isolate or drill down members to required dimension level. Sort data with advanced sorting options.
Create ad-hoc calculated measures, named sets, and use many more analysis options to shape your analysis and results that
will make a point. Well, you can play for a long time with this analysis and discover data you were probably not aware of,
but this time we need to make some quick analysis that will look presentable and focus on the main business aspects.
To check the visuals of your data just switch to Chart view and back, or continue to analyze while in chart view.
Options to dig for data are stunning. Finally, check your Report view to see the values and visuals at the same place
for end-users (the boss) to be happy with.
SQL/MDX Queries
To prepare data for dashboard tiles, except visual OLAP analysis, 'Kyubit BI' will offer you to create SQL, MDX, and
ODBC queries to get your data to be visualized. Simply query result will be the base for dashboard visualizations.
Result rows will become categories and columns will become series. Add any descriptions or share created queries
with your team if necessary.
Yes, your boss likes visual KPI indicators. Green, yellow, red, up or down, good or bad. Make him happy, add a few
KPI metrics to your dashboard, they look lovely to the management. Be careful - they should indicate how you stand
with your objectives, but often are misleading without the support of detailed analysis. Construct your KPIs with
values based on previously defined analyses and queries. Every time you open KPI, it will get fresh data for your
data sources and keep you informed about how you stand against goals, with the addition of nice line chart that
presents previous KPI values.
Dashboard design
Dashboard design and creating is the simplest thing that even a child could play with. It consists of dashboard
columns where you drag and drop various visualizations that you would like to see on your dashboard.
There are 18 visualizations of various chart types and metrics. After being dropped to the surface click and
select an existing analysis, query, or KPI to bind with a visualization. Now data is displayed on the dashboard
immediately and you start to believe you are going to make it. Select the number of dashboard columns and their
width, give the dashboard a title and description. Select initial Dashboard Filters (OLAP or SQL based) that slice
data on the whole dashboard. Everybody instantly understands pie and doughnut charts, though some Combo and
Bubble charts contain more complex data, while geography data is displayed using one of many Geo related charts.
Set appropriate users to be authorized to work with the dashboard as readers or contributors and your work is
almost done. Dashboard design actually calls you to interact, so users will click on dashboard analytic features
to drill down data to the point of interest or continue from a single dashboard visualization to full analysis view.
KPI Management and the purpose of Dashboard
Dashboard without metrics is hardly a dashboard, it's more like a report. You need metrics to tell you instantly
when dashboard is opened that certain things are going or not going according to expectations. It is like driving a
car and looking how much fuel you have and what speed you’re driving at. Dashboard is not a deep insight to your business,
it is a quick peek at your business objectives, but great dashboard software has ability to give you a deep analysis insight
when requested. So if a chart is telling you that sales last month were lower than you expected, you cannot react until you
find out the reasons. So you will click on the chart and select to continue with full analysis in a separate window that
will give you full blown analysis environment to continue drilling the data from the chart to various dimensions until you
get clear about why sales have dropped. Imagine that in an everyday business environment you have more than a few applications
and databases where a company has all kinds of data records that are relevant to the business. A manager should not jump
around from one system to another to get understanding of what is going on. You should prepare him a dashboard, a place where
all important indicators and values are on one screen. Furthermore, you can have one high level dashboard with the most
important data, while 'daughter' dashboards are opened on request for a specific segment of business. KPI is defined by values
defining what is success and what is failure in the sense of achieving defined goals. Values for both limits could be defined
from values that come from analysis and SQL/MDX queries which make them dynamic and by smart engineering they could last long
as a trusted source of information on your business score. Multiple KPIs could be organized to Scorecards, which is simply a
list of more than one KPI with its current status and attributes.
Subscribe
If the data on a dashboard is important to you to receive at time intervals,
fill in a subscription form and receive emails with the dashboard and fresh data from your data sources at scheduled times.
Even create conditional subscriptions that will send the dashboard only if one of its KPIs has required status.
People will collaborate on dashboard data by adding comments and starting dashboard related discussions.
Share and Collaborate
From the moment you send a link to other authorized users and they start to collaborate and discuss all kinds of
business related topics that are finally visible at a click in their browsers you will certainly find this tool a
powerful and analytic weapon in your battle to discover and share all data from your database. In most business
organizations there is huge amount of data still waiting to be discovered and bringing significance to your business
decisions that previously had low visibility. Having a dashboard where you can click on a pie chart and drill down a
pie segment to other dimensions such as time, product, customers, etc. is a fantastic opportunity for managers and
analysts who make business decisions on a daily basis, as they have a garden of their own business truths in front of them.
The ability for end users to comment on the BI content will attract users to see and make their own point and contribute to
the discussion. If KPI has failed status, opinions on a related subject are certainly going to flow and you do not want to miss
the debate and insights.
The dashboard is prepared and you send a link to other folks to see the results of your hard working opus.
Playing with dashboard slicers is the first interactive thing they are going to love. The boss has already exported your
dashboard to PDF, printing out data to show to his boss. He is satisfied that you have sacrificed your weekend for the
company.
Don't miss your chance to make your business data have a face that will smile or wrinkle :)
Author
Kresimir Korovljevic
Software architect on numerous projects related to Microsoft technologies. Development and promotion of Business Intelligence software tools is primary focus and driving force.