Choose your language:
MTN South Africa partnered with TEKsystems® Analytics Services to correct performance and usability issues associated with a failed implementation of Oracle Business Intelligence Applications (OBIA). Our remediation services allowed MTN executives to salvage a significant investment in Oracle applications and derive real value from the reports being generated to drive overall business decisions.
MTN South Africa is part of MTN Group, a global telecommunications company with more than 150 million subscribers across 21 countries in Africa and the Middle East. MTN South Africa delivers communications offerings and solutions to more than 20 million customers in South Africa. MTN has partnered with TEKsystems since 2013.
Business intelligence (BI) is a powerful asset that allows organizations to gain a better understanding of the overall health of their business and make timelier, more strategic business decisions. BI offers direct insight into current market conditions, consumer spending habits and overall business efficiencies, which organizations can subsequently harness to identify areas for revenue growth, cost-saving opportunities and the current successes and shortcomings of their operations.
For example, if a telecommunications company released a new phone to the market, they could review BI reports of real-time sales figures from the days or weeks following the release to measure demand for the phone. If sales were less than desired, the organization could almost immediately map a new marketing strategy to promote the phone and remedy the loss in sales without having to wait for quarterly revenue reports.
Leveraging the vital insight BI provides offers invaluable business benefits, though compiling it into a usable format can be challenging. With the vast amount of data available today, businesses must rely on complex and mature BI tools to retrieve and compile meaningful data from disparate sources, analyze it, and synthesize relevant data into comprehensive reports that can be easily digested and turned into actionable information.
The demand for BI has continued to increase in recent years as tools become more widely available, growing the industry to $14.4 billion in 2013, up from $13.3 billion in 2012.1 More organizations are looking to BI as a means to remain informed and tactical in an increasingly competitive marketplace. In fact, according to a recent Gartner report, chief information officers globally rank BI and analytics as their top priority in 2014.2
MTN South Africa was looking to gather greater insight on the effectiveness of their marketing campaigns and the overall health of their business by integrating BI applications into their enterprise IT environment. BI would allow MTN to run a variety of different reports that would provide concrete data to inform better business decisions to drive overall organizational performance.
Notably, MTN’s Marketing Department wanted to better understand the impact of promotions on their sales and how new products were being received by the marketplace. The chief marketing officer wanted to quickly gather and review sales data to determine which product promotions were generating business and should continue to run versus which should be suspended or discontinued. Additionally, better understanding consumer spending trends and behaviors would help guide the company’s overarching marketing strategy.
To achieve this level of clarity, MTN invested significantly in OBIA software and licenses so they could ultimately embed analytics throughout their business domains (e.g., sales, marketing and human resources). They would first implement the new BI system within their Marketing Department.
Given the complexity of Oracle’s BI products, MTN partnered with a third-party vendor to install OBIA; however, after being installed, the platform was impaired by numerous performance and usability issues rendering it useless to MTN business users. Issues included:
MTN executives were among the approximately 150 users regularly encountering system issues. Facing a slow and faulty platform riddled with inaccurate data, executives began to doubt the overall credibility of the system and their decision to further implement OBIA throughout their organization.
Over the next two years, their original third-party vendor attempted to remedy the extensive system issues to no avail. Frustrated by the shortcomings of the previous vendor, MTN considered discontinuing efforts to remedy the BI platform and moving toward an alternate BI product altogether.
However, while attending an industry BI conference in South Africa, MTN personnel had an opportunity to discuss the performance and usability issues plaguing their current BI system with TEKsystems’ South African Global Services® director. Based on their conversation, MTN learned of our extensive past remediation work helping other organizations facing similar BI roadblocks.
MTN executives, as well as Oracle, hoped that partnering with TEKsystems would help restore MTN’s confidence in the platform and allow them to leverage the BI analytics they had originally sought to enhance their business. MTN asked us to evaluate their existing environment to identify problem areas stemming from the initial platform installation and then subsequently propose and implement a solution to stabilize and restore system functionality.
Having previously completed numerous remediation projects for organizations grappling with similar BI challenges, TEKsystems was very familiar with best practices for such an engagement. At MTN’s request, we would first complete a preliminary health check and assessment of their existing environment over the course of five weeks.
To gather the appropriate information needed to craft a remediation plan, we would interview MTN’s technical team to learn more about the specific technical issues hindering performance and the overall system landscape. We would also review previously prepared reports outlining what attempts had been made to remedy the system.
After completing the discovery process, we would review the root causes of the platform issues with MTN leadership and outline a customized remediation plan to fix the identified bottlenecks. Based on our analysis, we would provide six highly-skilled resources with deep understanding of OBIA to take corrective actions to decrease the time to complete incremental data loads, eliminate user login issues and remap products where necessary to ensure data accuracy. We proposed placing resources under an on-site/off-shore model to help limit costs for MTN and minimize the on-site oversight that would have otherwise been required.
All corrective remediation actions would first be implemented within a preproduction environment and then, after receiving signoff from MTN stakeholders, be moved to production to ensure no additional end-user disruptions. Given the scope and complexity of the remediation, we estimated it would take approximately three months for us to stabilize the BI system.
After two years of struggles with the previous vendor, MTN nearly gave up on the Oracle BI project. However, TEKsystems was able to successfully resolve MTN’s BI platform performance and usability issues within three months and restore MTN’s faith in the feasibility of an enterprise-wide move to Oracle BI. As part of our remediation solution, we were able to drastically reduce the duration of daily data loads, speed login times, correct BI architecture and topology—including OBIA hardware configuration—and eliminate data issues, all while staying within budget. All 150 users, including MTN executives, are now able to run timely and accurate marketing reports.
During our initial health check and assessment, our team identified several key platform issues that needed to be addressed and corrected—the first major problem being OBIA’s architecture. OBIA is enterprise-wide software, so there are many intricacies involved with installation, deployment and configuration. The previous vendor lacked experience and had not followed best practices for installing an enterprise product, and as a result, the platform’s topology was implemented improperly, which explained several of the system’s performance and usability issues. OBIA is a three-tiered architecture product with three distinct and separate layers—an ETL layer, a business analytics warehouse (BAW) layer and a presentation layer. The previous vendor incorrectly installed two tiers, ETL and BAW, within the same server, leaving them to compete for server bandwidth and causing the processing delays. We immediately identified this error and restructured the platform to run on three-tiered architecture, the proper structure for an enterprise-wide product. This correction helped drastically reduce login times to mere seconds compared to the 10 minutes it previously took some users.
Correcting the system architecture also helped to significantly improve daily data loads times. Previously, within the extraction layer where data was being pulled to generate the marketing reports, ETL load times were reaching upwards of 14 hours to complete. To address this fault, our team implemented three-tiered architecture, upgraded the server to a recommended higher configuration and moved some of the filtering, mapping and business logic to the database and source level per best practices. We also cleaned up data quality issues by mapping MTN products correctly in the system. Through these changes, we were able to reduce ETL data load times to less than five hours and improve the data quality issues within the reports so users could rely on them to make business decisions.
Similarly, on the presentation layer we identified filtration rules that had not been implemented correctly or were completely absent from the system. Because of the filter anomalies, reports being generated were scanning the entire database of millions of records instead of a small subset directly pertaining to the user’s request (e.g., data from the previous day and data from the previous month). Our team worked to implement new logic and filters at the appropriate levels within the system, ensuring the requested reports were populated quickly and included accurate data. In all, we were able to fix and validate more than 400 reports. The number of reports providing meaningful and accurate data continues to increase, allowing MTN’s Marketing Department to derive real value.
Seeing our efficiency during the remediation phase, MTN asked us to prepare a three-year architecture and business roadmap for their overall BI program. One of the first recommendations given by our team was to upgrade OBIA 18.104.22.168 to the newest version, OBIA 22.214.171.124, which would help eliminate future technical issues and provide more business- rich functionality, including enhanced performance, and shorter and faster incremental load times.
As a result of our successful remediation and upgrade work, MTN has committed to using TEKsystems Analytics Services for all their future BI customizations and implementations. Over the next several years, we will work with several additional MTN South Africa functional domains, including sales, human resources and procurement, to implement OBIA software to better inform strategic business decisions.
TEKsystems was able to deliver successful remediation work through:
1 Press Release: Gartner Says Worldwide Business Intelligence and Analytics Software Market Grew 8 Percent
in 2013; Published April 29, 2014
2 2014 CIO Agenda: A North America Perspective, Gartner