Microstrategy Overview

Editor's Notes

• #4: Over the preceding decade, most companies have deployed BI applications as departmental solutions, and have accumulated a large collection of disparate BI technologies as a result. Each distinct technology supported a distinct user population and database, within a well-defined “island of BI”. At first these islands of BI satisfied the early needs of the bsuiness, but this early succes in departmental deployment has sowed the seeds of new problems BUILD SLIDE Successful applications always expand. They gain more users, more analytic scope, and access more data. The second era of BI is hallmarked by BI applications which have expanded to the point where they are no longer isolated islands. Instead, they overlap in user populations, overlap in data access, and overlap in analytic domain. As a result, CIOs are now faced with an untenable situation. The enterprise is getting conflicting versions of the truth. The quickly growing business user population is distinctly unhappy about being forced to use multiple different BI tools. And the CIO is wrestling with an ever accelerating maintenance burden to keep all of these disparate systems from tripping over one another in terms of definitions, data access, security, administration, and operations. BUILD SLIDE The next era of business intelligence is one where a single BI architecture delivers a single version of the truth through a single user interface to all people in the enterprise. It can access all of the data, administer all of the users uniformly, eliminate the repetitive data access, reduce the administrative effort, and reduce the time to deploy new BI applications. Such a BI architecture must exhibit two dominant characteristics to fulfill this role. First, it must offer the full range of BI functionality available in the myriad of existing products so that it can adequately replace them. And second, it must be “Industrial-strength” so that companies can confidently deploy applications across the enterprise. • Era 1: Isolated Departmental Islands of BI Are an Initial Success Many companies deploy BI applications as departmental solutions, and in the process, have accumulated a large collection of disparate BI technologies as a result. Each distinct technology supported a specific user population and database, within a well-defined “island of BI.” At first, these islands of BI satisfied the initial needs of the business, but early success in departmental deployment sowed the seeds for new problems as the applications grew. • Era 2: Overlapping Disparate Islands of BI Have Become an Enterprise Liability Successful applications always expand. The second era of BI is hallmarked by BI applications that have expanded to the point where they are no longer isolated islands. Instead, they overlap in user populations, data access, and analytic coverage. As a result, organizations are now faced with an untenable situation. The enterprise is getting conflicting versions of the truth through the multiple disparate BI systems, and there is no way to harmonize them without an extraordinary ongoing manual effort of synchronization. Equally problematic is the fact that business users are forced to use many different BI tools depending on what data they want. • Era 3: Enterprise BI Standardization Delivers a Single Version of the Truth with Lowest Total Cost of Ownership The third era of business intelligence is one where a single BI architecture delivers one version of the truth through a single user interface to all people across the enterprise. It can access all of the data, administer all of the people, eliminate repetitive data access, reduce the administrative effort, and reduce the time to deploy new BI applications. This is not just a vision of Enterprise Business Intelligence. Many companies are successfully achieving this new era of business intelligence with MicroStrategy technology.
• #5: A BI system is composed of several key components. The BI platform sits between business users you see at the top of this slide who consume the information, and all of the data sources your see at the bottom. The role of the BI platform is to interpret and harmonize all of the complex data across all of the different databases so that users and IT personnel can design reports and interact with data without requiring that they have any knowledge of the underlying technical complexities and inconsistencies that might exist in the databases. The BI platform provides a user interface to business people with which they can interact with their company’s data in any of 5 different ways – which we call the 5 Styles of BI. The five styles of BI are listed in the diagram and include: Reporting, Dashboards, OLAP Analysis, Advanced or Ad-hoc Analysis, and Pro-active Alerting. Many BI tools support one or two of these styles of BI within the same architecture, but with the MicroStrategy architecture, you can get all 5 Styles of BI within the exact same user interface and all built on the exact same building blocks (called metadata). And at the right you see that the real value of a BI platform is that it allows companies to create and deploy very many BI applications composed of reports, dashboards, and other analytic workflow… some of which are created by IT professionals, some are created by analysts, but ideally, the majority are created by average business people in the normal process of doing their various jobs.
• #6: Over the last 10 years or so, BI has evolved from simple reporting and analysis to 5 more concrete styles of BI. These styles came about because of an increase in the number and range of users who are accessing data. The requirement for more interactivity with data has also increased along with the need for more advanced analytics. Out of these requirements, the 5 Styles of BI were created. The 5 styles include…(quickly walk through each style).
• #7: The graph shows how every style is placed based on number of user and user interactivity
• #8: Show offline example: Claims Dashboard, Ca Dashboards often provide at-a-glance views of KPIs (key performance indicators) relevant to a particular objective or business process (e.g. sales, marketing, human resources, or production).[1] The term dashboard originates from the automobiledashboard where drivers monitor the major functions at a glance via the instrument clusterlifornia Population, Category Manager Dashboard. Dashboard are interactive
• #9: Enterprise reports are the very detailed reports that are the mainstream of reporting applications today. These highly structured, densely populated, multi-page reports are designed to convey large amounts of operational performance data. These reports are often organized hierarchically and may use banding to make them more readable. The need to be Print-perfect so business people can take them to meetings. They need to be pixel-perfect so any format and be achieved as is especially the case with business reports. And they need to be page-perfect so that they are ready for high production printing such as with bills and statements. Show example pds Retail and Manufacturing (under Enterprise reporting folder)
• #10: High volume, high efficiency engine allows to distribute the information to many people In order to avoid spamming users with reports every day or every hour, alerts can be configured so they are notified whenever there is something that requires their immediate attention
• #11: OLAP stands for On Line Analytical Processing. OLAP allows business people to easily display specific slices of data by performing any of the common OLAP functions of pivoting, paging-by, sorting, filtering, and drilling. OLAP is the foundation of mainstream analysis in organizations today because it is simple for users to understand. Basic OLAP - Basic OLAP With Basic OLAP technology, people perform these functions on a limited subset of data called a cube. With other BI technologies, these cubes are static physical databases, whereas with MicroStrategy, they are dynamically generated multidimensional caches. You can see in the diagram at left that users can view many different slices business performance: In the top slice, the user is viewing Laptop computer revenue for 2010 and for all Stores BUILD SLIDE - Relational OLAP Analysis Another approach to OLAP is Relational OLAP. Relational OLAP doesn’t limit you to slicing and dicing within a cube. Instead you are able to access information anywhere within your data warehouse, from the highest levels of data summarization, down to the lowest layers of transactional detail. Ideally, BI vendors would offer both approaches to OLAP analysis. Differentiator – Enterprise Class ROLAP Engine The Best In Class Approach Is: ROLAP architecture with sophisticated caching. Advanced ROLAP engine which allows people to drill anywhere. ROLAP engine that is optimized for each relational database.   (++) Which Is Important Because: Systems don’t fail or slow down even during peak usage. Any question can be answered at any time.   Other Alternatives In the Market Are: Some other tools store data in proprietary data storage mechanisms. Other ROLAP tools don’t have database-optimized ROLAP engines or caching engines. Other tools only allow you to drill within pre-defined subsets of your data warehouse, or along pre-defined drill paths.   (--) Which Are Less Optimal Because: Proprietary data storage mechanisms tend to not scale as well as relational databases. Non-optimized ROLAP engines require more resources (CPU, memory, network). Not all questions can be answered at any time by the business user. Additional Notes for Reference: The advantages of Relational OLAP are mostly obvious: People can investigate the data freely… without being constrained to just the data contained within a cube People can “surf” anywhere in the data warehouse… without the need to create new ad-hoc reports Normal business people can serve themselves with data ... without requiring assistance from IT Summary Having heard this, it should probably make you wonder why other vendors’ technology is not also based on Relational OLAP. The simple reason is that Relational OLAP technology proved to be very difficult for software vendors to perfect. It took MicroStrategy a decade of concerted work with customers, theoretical work, and many patents to produce a highly scalable, high performance Relational OLAP technology. Other companies spent this past decade investing in the easier cube-centered technology instead, and are just now realizing that customers’ requirements are rapidly evolving to need Relational OLAP capabilities. BROCHURE TEXT MicroStrategy provides both Basic OLAP analyses for simple analytical needs, as well as full Relational OLAP for more freeform analysis and more extensive investigation. Both forms of OLAP allow business people to easily display specific slices of data by performing any of the common OLAP functions of pivoting, paging-by, sorting, filtering, and drilling. MICROSTRATEGY DELIVERS ALL ANALYSIS REQUIREMENTS – FROM SLICE-AND-DICE TO INVESTIGATIVE ANALYSIS Providing Simple Speed-of-Thought OLAP Analysis to Business Users | MicroStrategy delivers quick insights to business users via Online Analytical Processing (OLAP). OLAP is the means by which a user examines data across various dimensions and hierarchies. With MicroStrategy, business users look for patterns, trends, and exceptions and intuitively change their views of the data through simple slice-and-dice analysis. MicroStrategy’s OLAP functions allow business users to understand and predict business events without being trained on analytic techniques. Open Up the Entire Data Warehouse with Relational OLAP Analysis | MicroStrategy uniquely delivers complete investigative analysis. MicroStrategy’s renowned relational OLAP (ROLAP) capabilities open up the entire depth and breadth of the data warehouse for full investigative drilling, down to the transaction-level of detail as necessary. With MicroStrategy’s “drill anywhere” capability, business users can surf anywhere in the data warehouse to see new sets of data, reducing the need for formal report creation or the help of IT personnel, and improving user comprehension of data.
• #12: The first is Data Discovery. And a good way to think about Data Discovery is Self Service 2.0. Let me explain. Slide Data discovery is an emerging style of BI that empowers Business people to conduct their own analysis without help from IT. For business people it is a brand new BI experience from going from data to Business insight.   You start with data. That data can be IT managed data from database or it can be your own data in spreadsheets. It’s up to you.   <build> You start by familiarizing yourself with the data. You can look into the data and begin to understand any relationships.   <build> Next, you can apply visualizations to that data to help you discern patterns, problems, and trends.   <build> You can apply your own analysis to this by filtering the data, creating new calculations, and adding views of the data.   <build> But truthfully, the data discovery process is not so linear. At any time you can add new visualizations, add new combinations of data, and refine your data analysis.   <build> Until you arrive at the business insight.   Data Discovery is an exploratory and iterative process. That exploratory process is highly visual. A good way to describe it is Visual OLAP. It’s a better way for many people to understand data than the classic OLAP tabular format. And business people can do all this on their own without help from IT. At the same time, IT can make sure this can securely deploy at scale and with high performance.
• #13: This is an emerging trand pushed by bigdata, it needs a data scientist
• #14: There are three types of roles within the analytics sphere. There are those who produce the analytics versus those who consume them. This provides to be a challenge for those in a corporation. This is because the data scientists utilize complex and technical skills to create analytics for the consumers. However, while business users and consumers want to gain the insight, they want to avoid the technical jargon and complexity that often comes with such data analytics.
• #15: Show a mobile app
• #17: The platform provides all 5 Styles of BI uniformly through a wide variety of user interfaces, including mobile phones, tablets, Web browsers, Enterprise Portals, Microsoft Office, E-mail, Microsoft Windows workstations, network printers, and file servers The integrated backplane is the heart of the MicroStrategy architecture. It provides the common services of metadata, prompt generation, scheduling, in-memory data, shared caching, security, user management, query generation, query governing, and administration. More importantly, it is the core engine which supports each of the 5 Styles of BI as plug-n-play “service modules” that can be mixed and matched in any combination. These modules include MicroStrategy Report Services for pixel perfect reporting, Mobile apps, dashboards and scorecards, MicroStrategy OLAP Services for speed-of-thought slicing and dicing, MicroStrategy Distribution Services for alerting proactive notification functionality, and MicroStrategy Transaction Services for data input and write-back capabilities. MicroStrategy ROLAP and analytics are always an integral part of the core engine. MicroStrategy customers can add various “service modules” to the core engine to incrementally extend its functionality. Each service module is designed to build on one another, adding new functionality to each other when used in combination. This means that a customer can add the OLAP Services module to the engine, and all previously-built grid reports automatically become Intelligent Cube™ reports, and inherit a wide range of new functionality. In addition, a customer can add the Report Services module, and be able to re-use all the previously-built grid and cube reports as datasets for the new pixel-perfect document. Furthermore, the Transaction Services module can be added to enable data input and write-back functionality on Report Services documents and dashboards. And finally, it means that a customer can add the Distribution Services module, and immediately use any grid reports, intelligent cube reports, and document reports as the basis for e-mail alerts and proactive notifications.
• #18: An organic, platform of 17 distinct products that dynamically assemble, and re-assemble these objects in response to any business question or user action. Every MicroStrategy BI application requires at a minimum: 1. Development products to build business abstraction objects in the metadata: MicroStrategy Architect and Desktop. 2. Core server to manage users, security, database communications, report formatting, and distribution: MicroStrategy Intelligence Server. 3. Interface products to run reports and manipulate the results for investigative analysis: MicroStrategy Mobile, Web, Office, or Distribution Services
• #24: With in-memory ROLAP, some selected portions of the global virtual cube are populated with actual data as ROLAP Cubes (Intelligent Cube); as you see presented here in the drawing. BUILD Because the in-memory ROLAP Cubes operate below the User Security and Prompting layers, they are able to serve a very wide variety of requests and user profiles. So ROLAP Cubes have flexible query characteristics of databases, but because the ROLAP Cubes reside in main memory, they have the performance characteristics of caches. On the performance graph on the right you can see the addition of a new performance curve for the in-memory queries. These in-memory queries have replaced a large number of database queries and delivered their data with much faster response times. Introduced with MicroStrategy 9 last year In-Memory Cubes can reduce the computational distance for many reports. In-Memory Cubes have query characteristics like databases, but because the Cubes reside in main memory, they have the performance characteristics of caches. This technology is designed to overcome the limitations of Caching, opening up the technology to a wider range of reports. In-Memory Cubes provide 1-3s wait time to a range of 40%-60% of reports in a BI application. These superior performance characteristics make In-Memory BI the best option to improve performance significantly
• #25: Now here’s the huge effect of the In-memory operation on the Database Workload. These huge spikes in user activity show up where the volume of report requests overloads both the Intelligence Server and the Database. In-memory adds some batch processing to the end of the ETL window as the Cubes are created with the updated data… …but the workload is greatly reduced as huge numbers of reports can now hit the Cubes instead of the database.
• #34: Beginning in 1996, MicroStrategy embarked on a massive software development effort to re-engineer its highly scalable architecture so that it could support all 5 Styles of Business Intelligence. By 2000, MicroStrategy delivered MicroStrategy 7 which laid the foundation for this vision, and explicitly supported the two most difficult styles of BI: 1) Advanced Analysis, and 2) Alerting. By mid-2002, MicroStrategy added In-Memory to the platform and further enhanced this in 2009. In 2003, MicroStrategy added Enterprise Reporting and Scorecards & Dashboards. Data Mining was added in 2005. MicroStrategy released its Mobile App for Apple iOS phones and tablets in 2010 and for Google Android devices in 2011.
• #37: PRESENTATION NOTES: In order to achieve development efficiencies, a modeled environment breaks the essential components of BI applications into a metadata. The metadata is reusable and object orientated. From a small number of base objects, you can create all the reports you need. Introduce the concept of a business model and hence, logical model. Introduce the concept of a layered MD – and Say how in a layered MD, the business model is abstracted into a logical model of three layers (schema, business abstraction and reporting layers) In short, the logical Model is a representation of the Business Model.
• #38: Data abstraction objects map the physical data model of the data source into a logical business model using business terms and rules. Data abstraction objects serve as the building blocks for business abstraction objects and, ultimately, for all reporting, analysis, and monitoring
• #39: Explain each Key component briefly and highlight the reusability of these objects Table – nothing but a logical presentation of the physical database table Fact – A fact is a table column, or an expression combining multiple columns, that contains quantitative raw or aggregated values such as amounts, counts, quantities, forecasts, and other measurable business calculations Attribute – Attributes are defined by a column or columns in one or more tables, and provide meaningful qualitative context to facts. Attributes are commonly displayed on reports. They also serve to set the aggregation levelfor calculations in a report, and for filtering the data that is retrieved from the data warehouse. Examples of attributes are regions, stores, customers, and months. Dimensions or Hierarchies logical groupings of the attributes. Talk about the Time Hierarchy because that Hierarchy is the easiest to understand and is common to all organizations. E.g. Year  Quarter  Month is a logical grouping that form the Time dimension. Also, highlight how Year and Quarter form a Parent child relationship and how Quarter and Month are also the same. e.g. Year 2006  4 Quarters (Q1-Q4 2006)  Each Quarter has three months (Q1 2006  Jan, Feb, Mar 2006) etc. Hierarchies: A hierarchy is a logical group of attributes that defines a relationship between them. A hierarchy describes an ordered path through the attributes from a summarized level to more detailed levels. Hierarchies are used to browse data, and provide a navigation path when drilling from one attribute to another on a report. In MicroStrategy, hierarchies also serve as placeholders for an attribute in a report, with the specific attribute being determined at run-time. There are two types of hierarchies: • System Hierarchy The system hierarchy is a direct map of the logical table structure stored in the metadata, and shows the true relationships between all attributes in the reporting application. There is only one system hierarchy for each MicroStrategy application. • User Hierarchies MicroStrategy allows a greater range of hierarchical paths than those found in the physical database. User hierarchies define relationships between attributes based on the business rules of the reporting application, and they guide user analysis or provide shortcuts to common analytic workflows. Attributes can be reused in different user hierarchies. Transformations – they are the most complex and are used to provided time period comparisons of facts. E.g. compare this year revenue to last year revenue. YTD or MTD etc. Transformations are reusable business abstraction objects that are embedded primarily in metrics for time comparison calculations. They automatically display information for changing dates and time periods without constantly needing to redefine metadata objects. For example, many reports display current period information along with prior-period and period-to-date information. In this example, transformations handle the variable time periods automatically regardless of when the reports are executed within the same object definitions. Two types of transformations are supported in MicroStrategy: • One-to-one transformations compare one period to another, for example, “Last Month” and “This Month Last Year”. • Many-to-many transformations define period-to-date and last-in-period metrics, for example “Month-toDate” or “Last Four Weeks”.
• #40: Business abstraction objects define the business terms and rules. They serve as the building blocks for other business abstraction objects and for all reports, analyses, and documents.
• #41: Metrics: Metrics represent business measures and key performance indicators. Metrics consist of facts, attributes, or other metrics combined with a formulaic calculation. This formula may include aggregation functions such as “Avg” , “Sum” or “Count,” arithmetic operators (+, -, *, /) or other advanced functions (e.g. data mining, financial, mathematical, OLAP, statistical). A metric can contain 3 additional properties: • Dimensionality Metric Dimensionality determines the level of aggregation of a metric. The default aggregation level is based on the attributes contained in a given report. The aggregation level of a metric can be changed to force aggregation at a specified level. This flexibility is especially useful if business metrics need to be calculated at the same level regardless of the level of reports that use them. • Filter Metric Filters determine any restrictions on the data used in the calculation. A metric filter may be defined as an embedded component of the metric or as a link to existing filters, maximizing object reuse, and reducing administration. • Transformation Transformations may be associated with a metric to enable time series analysis as described in the next section. Filters: A filter specifies the conditions that data must meet to be included in a report or metric. Filters limit the data that is extracted from the data source, and focus the data presented to the end user. A filter is essentially a logical expression that combines any metadata object, mathematical and comparison operators, and constant values and lists. Filter expressions can be combined using Boolean operators and nested to create sophisticated conditional clauses. For example, to retrieve data for New York and San Francisco for the fourth quarter of 2012, the filter would be: (City=’New York’ OR City=’San Francisco’) and Quarter=’20125Q4’. Templates: A template specifies the layout and formatting of the data on a report. Attributes, metrics, custom groups, consolidations, and prompts can be placed in rows, columns, or pages on the template grid. There is no theoretical limit to the number of objects that can be placed on a template. Besides being able to define the presentation characteristics, such as font, color, alignment, and number formats, additional report characteristics such as subtotaling, metric thresholds, and graph properties are also stored in templates. Prompts: Run-time prompts provide report input parameters that control most aspects of a report, and give the user significant ad hoc reporting capabilities. Prompts enable the personalization of reports from a single report definition, reducing the number of objects stored in the metadata repository. There are four types of prompts in MicroStrategy: • Filter prompts provide filtering criteria at runtime, and qualify the data set from the data warehouse that will be included in the report results. • Object prompts let users choose which attributes and metrics to include on a report at run-time. They can also be used to build metadata objects, particularly metrics, providing enormous flexibility from a single report definition. • Value prompts allow the input of alphanumeric and date values that specify constants for filtering criteria and metric calculations. • Level prompts dynamically control the aggregation level (dimensionality) of metrics.
• #42: This represents tremendous savings in terms of development and maintenance overhead, because I can define an object only once, and reuse it as many times as I want For instance if I create a filter, I may use this filter on: So the benefits are twofold: First I save on development overhead since I don’t have to build several copies of the same object Second, If I ever have to make changes to this filtering criteria, I only need to change/update objects in one place only! So I save on maintenance overhead too.
• #44: A quick show throug the 300+ out of the box Microstrategy function
• #49: A possible interaction between business users and IT could be through the use of VA by the business users as a prototype instrument, in an agile development way IT collect this prototypes and develop dashboards that will be published to the crowd. VA dashboard could be even published directly
• #50: Probably the best statistical graphic ever drawn, this map by Charles Joseph Minard portrays the losses suffered by Napoleon's army in the Russian campaign of 1812. Beginning at the Polish-Russian border, the thick band shows the size of the army at each position. The path of Napoleon's retreat from Moscow in the bitterly cold winter is depicted by the dark lower band, which is tied to temperature and time scales. Exquisitely printed in two colors on fine archival