Essential Strategies Of BI Implementation

A successful Business Intelligence Solution needs a strategic execution of implementation process.I have divided the overall process into:

Business Intelligence Modules:-
Although the core need of Business Intelligence Platform remains uniform across all organization, the actual requirements vary from department to department and from organization to organization. Some of the BI products are one-business vertical specific and they may not be suitable for another business domain. Mostly, BI vendors offer BI functionality to deliver a solution to the common sets business requirement and they remain common across all organizations and implementations

In every BI implementation, the selection of right Business Intelligence Platforms is the key step for better success and higher ROI (return on investment) from BI initiative and implementation. Bunches of BI Platforms are available in the market and each of them is best in its own area. Apache Hadoop with Teradata is best solution for handling the Big Data and it may not much appropriate if there is smaller volume of business data. The first and foremost parameter to consider in BI initiative is the money budgeted for the BI project. The selection process of BI Platform including ETL Platforms and EDW should consider the allocated budget amount too.

These days, BI Platform from the same vendors is available in two different favors namely completely owned propriety software bundle and BI Solution offered as SAAS. SAAS based BI solution is generally offered in Cloud-SAAS mode making it suitable for larger and smaller enterprises. BI Platform offered as Cloud-SAAS is always suitable for economical and time-constraint implementation as the solution needs only the operational and implementation cost and does not need any capital expenditure to maintain IT-infrastructure. It saves time through its ready-to-use BI Platform availability.

In theory and the best practices, the need of different component or modules of BI Platform starts purely at pre-formatted enterprise reports and ends either with Business Analytics or Business Cockpit. This trend makes selection process very difficult and strict as it is always better idea to use maximum possible components or modules (Reports, Analytics, Dashboard etc) from one platform if the initiative is being considered from the scratch. Otherwise, suitable component from any one of the BI vendors can be considered if it is falling under the economical boundary created by the allocated budget.

BI Module Selection-Process and Methodology:-
If need of BI solution is pre-formatted report oriented, it is important to consider the robustness, the support for custom-program and API used to build the enterprise reporting framework or platform. Approximately, only 25-30 percent of ER reports are complex in layout and presentation and they need custom programs or coding to get the desired output and layout. The rest of the layouts and presentation-format are always served by default functionality. The lists of the features to be considered in this type selection process are: –

  1. Flexibility of Designer  & web-based User-Designer
  2. Server-Client Architecture.
  3. Heterogeneous support for data sources i.e. the support for all widely used databases, connectivity protocols like ODBC, JDBC, Web Service, MDX, and OLEDB.
  4. Support of external or custom program.
  5. Data-Grouping based on functional level security model.
  6. Report output version and archiving.
  7. Data export into PPT, PDF, Excel, CSV.
  8. Scheduling of report and delivery of report outputs through externalized channels like email and in future it could be the updates to Social Networking Sites like Faceboo, Twitter etc.
  9. Support for Social Networks Data and delivery of outputs to mobile devices like Smart-phone, Tablets etc.
  10. Offline analysis of report contents.
  11. Concurrent execution and viewing of the same report-output by multiple users
  12. API to include a piece of report inside a web-portal or Customized Dashboard interface.
  13. Roadmap of product for the support of Big Data, Location or Geographical Intelligence.

In 90 % of this kind of scenario, Business Cockpit or Dashboard is always considered and KPI(Key Performance Indicators) are presented through it. Initially, its user-adoption rate may remain low but it will go up very quickly as any business manager or executive always want to have bird-eye-view of overall business in one screen and it is leveraged by Dashboard or Business Cockpit and Scorecard.

Conclusively, the selected BI Platform should have strong support for pre-formatted reports and all standard features of BI Platform to make it suitable to cater the solution to the need of pre-formatted reports.

When the expectation of BI Solution is also an analysis of business data through Business Analytics or Self/Predictive Analytics, selected BI Platform should have Business Analytics or similar functionality to support the analysis of Business Data. The some of key parameters to consider on selection process are:

  1. Business vertical dominance of the platform in market.
  2. Performance Case Study when there is chance to getting more than 100K records in one Analytics Profile or Design.
  3. Availability of pre-formatted layouts and scope of on-fly customization of look-and-feel.
  4. Architectural process of Business Data and Business logics
  5. Concurrent access of same profile/design my multiple users
  6. Offline analysis capability
  7. Integration and customization supports
  8. Technology used and availability of the experts in the market

Almost all BI vendors are offering this kind of solution (Business Analytics) either in 100% proprietary or Cloud-SAAS model. Here also, the same thumb rule is used, if budget and time frame are not the constraint then BI platform can be deployed on Cloud-PASS/Cloud-IAAS to get privately held Cloud-SAAS solution. Otherwise, Cloud-SAAS offering from selected BI vendor can be considered after checking the availability of the offering and second option is always good as it falls under best practices of BI Solution.

OSI BI Platforms are also evolving very fast and successfully in BI market and the same catch will come in the selected part of Indian BI market. So, it is also not a bad idea to consider open source BI platform like BIRT, Pentaho, and Logixml during the selection process. The TCO of the BI solution designed and leveraged through open source BI platform will always be lesser and the capital expenditure of owning the BI platform can be utilized to the operational cost. Its actual cost is only for the consulting to be taken from the Freelance BI Consultant/experts or experts from the consulting firm.

Finally, to conclude the process of selecting the BI platform, it is a general practice to review the functionality and features through POC (Proof Of Concept) and it is offered by every BI vendor. The POC is offered as free of cost under their pre-sales strategy. Some of vendor has a practice of charging some consulting fee even for the POC could be to maintain its brand value, to get execution cost and to make their pre-sales strategy stronger and standard in the market.

Best Practices- Business Intelligence:-
Once the right BI Platform is finalized under the guidance of organizational rules and policies, the implementation process can be initiated.  Business Requirements Analysis stage is the vital and critical phase in whole SDLC of BI Solution. It is always necessary to capture the complete scope and details of business requirements for which solution is expected from the selected BI Platform. They can be collected unit wise through an individual, one-to-one interaction with users, business manager and CXO and at end of analysis they can be merged together to create the completely holistic view of the requirements. The requirements and expected Business Value Addition of BI solution can always be captured in standardized documented format and they are called FDS (Functional Specification Details). To create the details of this document any standard approaches like Use-Case can be used. FDS may also include sample design of reports, Business Analytics and Dashboard, business case studies and Solution Prototype to refer.

To make a BI implementation a great success and supper hit, it is necessary to design and create the project related documents with minute details and it comes as part of Risk Management Strategy of the project. The lists of document to be prepared under general guidelines of BI project are: –

  1. Project Estimate
  2. SRS (Software Requirement Specifications)
  3. FDS (Functional Specifications Detail
  4. Solution Prototype
  5. DDS (Detailed Design Specification
  6. Test Plan
  7. CR & Review Documents

The selection process of Executive Dashboard and Scorecard should consider easiness, compact and usefulness of the exposed functionality at real-time and it should not be considered from the technology aspect as most its users are always high-level business executives. The primary selection parameters for Dashboard or Scorecard can be Innovative Visualizations, Automatic Intra-Correlation of Visuals, and High Degree of configurations, Real-Time Alerts, responsive interactivity and banners or Ticker.

Irrespective of size of the organization and number of the reports, analytics and Dashboard KPI or items, it is always a better strategy to create fully configured BI Platforms. Three different software environments may go in place for the development, test and production of BI solution.

Modularization of delivery of BI Solution is always a good idea in any BI implementation and it reduce the risk of getting request for bigger enhancement or changes at later stages of SDLC.  The conceptualization of BI solution using the selected BI Platform to cater the solution for the gathered requirements is a great art it can be carried out by an experienced and an innovative expert to get better business value addition. It can include the details on designing standards, standards for EDW and Business Analytics, Business Cockpits, integration, technical testing plan and UAT/UWT plans.

The proposed BI Solution can be broken up into different modules. If the BI initiative is from the scratch, the designing of logical and physical database for the reporting and add-on solution can be considered as first modularized unit of overall phases. Modularization helps in getting the user feedback during the first module itself and it gives the room in avoiding repetitiveness of already occurred functional and other mismatch or errors in other up-coming modules. This practice lowers TCO and gives the higher ROI with higher delivery of Business Value Addition. Each suggestion or change request suggested by end users, business users, technical and functional managers and BI Architect/Consultants during UAT and UWT can be analyzed and considered to make the better practices for next modularized unit of the project and other future delivery to cater the best BI solution in industry.

Author: Amrit Chhetri,


Tagged: , , , ,

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: