Self-service BI is the one of the top trends in the market. The speed of business has dramatically increased. As a result, there is a high demand for quick decision making based on data. Business users want to connect to their data, immediately discover it, and extract business value from the data. They don't want to ask IT and wait a long time for their request to be processed. Moreover, they want a friendly interface that is intuitive and provides simple drag-and-drop options in order to slice and dice data. This is one of the latest trends in the analytical market. As a result, MicroStrategy have developed their own product for self-service data discovery analytics that can add value to the organization by decentralizing BI.
MicroStrategy Desktop puts the power of visual data discovery in our hands. It is designed for self-service analytics and gives us the freedom to analyze business data effortlessly with an intuitive, friendly interface. As a result, business users can easily connect a number of various data sources such as enterprise data warehouses, flat files, spreadsheets, social media data, and so on. Moreover, they can clean and transform data for better analysis.
This chapter will cover the following points:
- Connecting to various data sources
- Building ad hoc schema
- Data visualization
- Using D3 visualization
- Publishing
Let's look closely at MicroStrategy Desktop. In order to download the installation files, we should go to the MicroStrategy website (http://www.microstrategy.com/us/free-trial) and choose MicroStrategy Desktop. If we don't have the license key, we can obtain a 30-day trial period. It is enough for a proof of concept project or just for playing with the tool and getting a first impression of it.
There are two versions of Desktop:
- Windows 64-bit
- Mac
This gives us the opportunity to run Desktop even on Mac OS. There are many companies that prefer to use Mac instead of classical Windows. It is gives you much more flexibility. In our case, we use the Windows environment, where we run MicroStrategy Developer.
Let's install MicroStrategy Desktop and learn it through several exercises. In order to install it, just run the installation file and it will do all the work.
These are the main features of MicroStrategy Desktop:
- Powerful Desktop analytics
- Easily connects and blends data from any source
- Cleanses and prepares data with built-in data wrangling
- Unparalleled data access
- Intuitive dashboard design
- Powerful predictive analytics
- Large data visualization library
As we mentioned before, MicroStrategy Desktop allows us to make BI decentralized. MicroStrategy Platform allows us to use both BI types, decentralized and centralized. See the following diagram in order to understand the difference between these two approaches:
Centralized BI usually is managed by the IT department. It should be a reliable platform that offers accurate, quality data through reports and dashboards. The source of all data is the enterprise data model, mapped to the MicroStrategy schema. All attributes and metrics should be conformed and there is a small chance of getting wrong data. On the other hand, we have decentralized BI, where everyone, who has enough skills, can connect any source of data, blend it, and find business insights. In the preceding figure, three numbers demonstrate possible scenarios:
- Connect any data source such as flat files, big data sources, NoSQL data stores, web services, transaction databases, and so on in order to quickly discover data or build dashboards.
- In the same way, users can connect trusted schema attributes and build dashboards with clean, accurate data. In addition, they can enrich data using any data sources.
- The final dashboard can be promoted to the enterprise platform. Before promoting any reports or dashboards, IT should review and approve it. In addition, users can just share their insights using e-mail.
Let's compare these approaches:
|
Ad hoc schema |
Schema |
Ad hoc schema connected to the EDW | |
|
Data |
Agnostic data source |
SQL or MDX sources |
Enrich schema data with agnostic data sources |
|
Result |
File with a localized security model |
Globally-applicable set of objects that leverage the unified security model |
File that hooks into the central object and security models |
|
Value |
Fast and dirty, speed to value for ad hoc analyses |
Single version of the truth |
Quick speed to value and ability to access the central object and security models |
In addition, we want to point out the pros and cons for classical schema and ad hoc schema. Usually, developers use MicroStrategy Architect in order to build schemas. It gives fine-grained control over database concepts, schema objects, and data governance. However, it takes a long time to build complex schemas. In terms of the ad hoc schema, it dramatically increases speed of development, allows us to prototype use cases, and validate data. Unfortunately, ad hoc schemas are disconnected from the single version of truth and security models for enhanced flexibility.
MicroStrategy allows the following workflow in order to handle all types of schema:
- Create a Prototype: Build an ad hoc schema using the Data Import tool to validate the concepts, the data, and the metric you want to build on top of your datasets.
- Make a decision: Decide with stakeholders whether this data and these objects are worth incorporating into your company's single version of the truth.
- Add to Modeled Schema or Use cube for Analysis.
Before we start to work with MicroStrategy Desktop, we also want to point out the fresh magical quadrant 2016 of Gartner for BI tools:
There is a full report here: https://www.microstrategy.com/us/go/gartner-magic-quadrant-16.
In Chapter 1, Getting Started with MicroStrategy we looked at the Gartner BI quadrant of 2015 where Enterprise BI, such as MicroStrategy and SAP BusinessObjects, with self-service data discovery BI ,such as Tableau and Microsoft, were in one leader quadrant. However, as we see in the leaders' quadrant 2016, there are only self-service BI tools, such as Tableau and Qlik. It means that there is a shift from Enterprise BI to self-service BI. However, it is not obvious. Anyway, lots of people still use Excel spreadsheets.
In order to work with self-service BI, technical skills are still required, especially with Qlik. As a result, all BI vendors try to offer something that can help users to easily work with data. For example, SAP offers SAP Lumira and MicroStrategy offers Desktop. Most of them have something in common: their interface, which is very similar to Tableau and in fact is becoming the standard of self-service BI.