Oracle OBIEE in Utilities

Introduction

An OBIEE Success Story: How a Regional Utility Created Visibility in Supply Chain provides an overview about a project utilizing both OBIEE and Business Intelligence Analytics products. The project’s goal was to provide timely data and reporting to Supply Chain for aiding in strategic decision making. The result was a reduction in overall operational costs, performance and productivity tracking, inventory management in partnership with business operations and the initiation of basic governance practices for the data within the Oracle E-Business Suite.

Background

The Regional Utility has been in operation for over 100 years and has 1,500 employees that serve over 500,000 customers in 24 counties.  The company has 3,700 approximately megawatts of generating capacity at 10 power plants located primarily along the Ohio River.

The Supply Chain group consists of 11 full time buyers and manages $337M in annual spend.  The Oracle E-Business Suite Purchasing modules have been implemented since 1999.

Business Case

Often, the first and largest challenge of any project is obtaining organizational support and funding.  In order to gain approvals, organizational management wants to clearly understand the benefits and return on investment (ROI).  Projects that have potential impact on the bottom line will have a much higher probability of getting funded.  To garner support for our business case, we performed cost benefit analysis to determine the ROI.

Supply Chain provides a great foundation for OBIEE because the opportunities for savings are great.  As a team, we justified our project through a combination of excess and obsolete inventory reduction targets.   These targets were achievable through active inventory monitoring, as well as event impact analysis.

Event impact analysis through simulation allows for proactive inventory planning and informed decision-making when planning for seasonal changes in inventory.  The data provided through this analysis leads to informed, strategic decisions, with factual historical data.  The ability to show historical data and predict future events in a visual format creates a powerful decision-making tool.

Project Planning & Goals

With any successful project clearly defined goals are crucial. This project, for example, had 3 main goals: 1) Reduce overall inventory costs, 2) Manage inventory more efficiently and 3) Track operational performance. With these goals in mind, we initiated a project using Project Management Institute (PMI) methodology. Our first step was to create a project charter, which serves as a statement of scope, outlines project success factors and identifies key project team members and organizational stakeholders.

It was through this scoping process that we quickly determined that our requirements were vague and subject to frequent change.  To mitigate this risk, we chose an agile delivery model by decomposing the metric development into two to three-week iterations. When we neared the end of each cycle, we would plan the next iteration by taking what we had learned from the previous iterations.  This is a form of PMI’s progressive elaboration, or rolling wave planning.  This approach allowed us to be flexible, to deliver more metrics than originally planned and to gain a high rate of user acceptance.

Challenges

Throughout our project, there were a number of obstacles we encountered, resulting in lessons from each of those.  One key lesson we learned was to pilot the module for the users early in the project timeline, so that users have a much easier time discussing processes and requirements when they can see the actual system which they will be using.  Piloting early helps to eliminate anxiety from the user group while solidifying requirements.

By keeping the project agile, we allowed users to see and touch the product early.  This allows the business to develop their requirements with each development iteration and makes for a better final product.

Data quality can also introduce roadblocks to any business intelligence project.  By remaining agile, we were able to shift focus to other metrics in the release while data issues were resolved.  Agile development requires that the project team communicate clearly and effectively in order to be successful.

A second lesson we learned was that face-to-face user reviews led to better adoption.  It allowed project resources to observe how power users expect to navigate through metrics.  This led to user-friendly metrics and increased the confidence of our power users.  Also, users could see changes adopted in development which led to a sense of ownership.

Lastly, being a lean organization, we face the obstacle of resource availability with every project.  By defining clear roles and commitments in the project planning phase, we were able to circumvent some of the issues that arise out of resource availability while simultaneously ensuring that our resources were performing their necessary tasks to keep the project on course.

Successes

Data quality can be a large obstacle for any business intelligence implementation, so this project allowed us the opportunity to revisit old processes.  We reviewed and cleaned up several data fields to ensure the integrity of our developed metrics, including:

  • Item Types
  • Purchasing Categories
  • Single Source Justification codes
  • Item Last Review Date
  • Item Status Codes
  • Item Lead Times.

Training is the key to getting users and suppliers to adopt the new technology.  Because the client has several locations, we found that using multiple platforms for training was a most successful approach.  We found that by having training at several points during the project, we were able to incorporate user feedback in subsequent releases.

We also developed a governance policy for ongoing data maintenance to ensure continuity in metrics going forward.

Setup & Configuration Leading Practices

Two foundational components to the agile rollout and immediate analytics value were Oracle Business Analytics Warehouse (OBAW) and Oracle Business Intelligence Applications (OBIA).  These pre-built complements eliminated the data modeling and extract design that contribute risk and expense to a traditional data warehouse solution and allowed the client to leverage Oracle created accelerators to jump directly into building tailored metrics on a weekly basis.

Oracle Business Analytics Warehouse (OBAW)

OBAW allowed the client to sidestep the discovery and design effort to create their own fact and dimension tables, which are core to a business analytics.  This complete repository has all required relationships and grain established for all entities that allowed the project team to build on this foundation and create tailored summary tables that combined several entities without the risk of changing grain or inaccurately joining entities.

The result of leveraging OBAW was a focus on the exact data from the warehouse they needed for their analytics, instead of spending time building consensus on entity definitions and validating foreign key relationships.

Oracle Business Intelligence Applications (OBIA)

OBIA filled the gap between the operational eBusiness suite source and the OBAW warehouse target.  OBIA provided the complete set of extract, transform and load (ETL) logic required to move operational data from eBusiness into the OBAW fact and dimension tables.  These ETL packages are pre-built in the industry leading PowerCenter tool by Informatica.  Leveraging these mappings provided two distinct opportunities for cost saving:

  • PowerCenter is a commodity skillset: Informatica PowerCenter is the standard for moving data between systems.  As a result, skilled resources are readily available at a low cost to extend OBIA extractors to suite specific needs.
  • Oracle built mappings expecting customer extensions: Oracle expects its analytics users to have requirements beyond what is provided out-of-the-box.  As a result, the OBIA mappings come preconfigured with generic expansion attributes. These attributes simply need to be re-named to make their content obvious and trace the attribute back to the proper eBusiness attribute following pre-built mappings to the same area.

OBIA enabled the project to quickly incorporate eBusiness customizations into the OBIA extract logic without impacting the all-important grain of the fact and dimension tables.  This also means the eBusiness customizations flowed easily into the data warehouse to be used in metrics and dashboards.

OBIEE Presentation Layer with Oracle Answers

The final key feature to the client’s setup was the use of Oracle Answers as part of the OBIEE presentation layer.  Oracle Answers is a web based report and metric builder that relies on drag and drop interaction to create metrics and reports within OBIEE.   While this tool does require an understanding of common data manipulation processes, such as group by, math functions and string manipulation, it does not require complex SQL or programming skills to build metrics quickly.  The project team worked with the business to identified team leads in each business area that would gain experience in modifying existing metrics to provide business unit self-service capabilities for most on-going tweaks to existing metrics.

Summary

In summary, we feel there are three key aspects to a successful implementation. First, as soon as you have gathered basic requirements, stand-up a pilot system to further refine your requirements and business process. This solves the “I don’t know what I want until I see it” issue.  Also, by keeping your project agile, you allow room to develop requirements along your project timeline. Second, ensure you have adequate time and money budgeted to support data cleanup efforts.  The business intelligence dashboards depend on quality data in order to be effective for decision making.  Lastly, ensure that you budget time for training and the creation of a mechanism to incorporate feedback into subsequent releases.

To learn more about AXIA’s services or to schedule a business consultation, contact us at 866-937-5550.