by Tara Swords
In 1967, when the IT industry was still in its nascent phase, a startup IT consulting company launched in France. That company, Sogeti, wouldn’t stay little for long; it eventually came to be known as Capgemini and grew into a global force with 140,000 employees and operations in more than 40 countries. Now, Capgemini is one of the largest consulting, technology, and outsourcing companies in the world.
Much of the company’s growth has happened through acquisition over the last 30 years. And while acquired companies were integrated into the fabric of Capgemini, they preserved their ability to operate as independent regional entities. The strategy worked well because it was based on the concept that the employees of the acquired companies knew their local customers and business climate better than anyone else could.
“It’s embedded in our DNA—a local entrepreneurship,” says Eric Jahn, finance transformation manager at Capgemini. But the idea of local autonomy has significantly evolved over the past 10 years. Today, many of the company’s leaders are driving toward a different vision: that of global shared services and processes as a way to increase automation and efficiency. One initiative, which the company introduced in 2008, was the Global Financial System (GFS)—a suite of applications that handles financial accounting and financial operational reporting. The GFS runs on Oracle E-Business Suite and generates reports through third-party systems and an in-house-developed application.
The company’s IT organization was making steady progress, bringing all countries online with the GFS. But early in the implementation, serious scalability issues were causing major delays in critical financial reporting. This added friction to the newly minted shared services process, complicating the needs of regional users and slowing the rollout of the system to remaining regions.
|Location: Paris, France
Industry: IT services
Revenue: €10.1 billion in 2013
Oracle products: Oracle Exadata Database Machine, Oracle E-Business Suite
|Length of tenure: 10 years
Education: MBA, State University of New York, Albany
Capgemini leadership reviewed the existing platform and identified a need for increased computing power to crunch the new streams of financial data flowing into the system. Leaders of finance IT started looking for a solution that could dramatically speed up reporting and ensure that the GFS could scale to users’ needs—driving new efficiency into its operations around the globe. Oracle Exadata delivered the performance Capgemini needed to expand the GFS to multiple countries.The Impact of Performance
For all of Capgemini’s global service lines and business units, project accounting is the most important function of requiring financial controls. The company employs thousands of consultants and organizes business by projects—as many as 100,000 active projects are live at any one time. The GFS exchanges data with an in-house system called N2K (“Need to Know”), which project managers use to keep tabs on project progress. This data shows which budgets and schedules are on track, monitoring how much time team members are contributing to projects and comparing actuals to forecasts. The goal of the GFS is to report revenue consistently and ensure that expenses are in line with budgets—as well as demonstrate to clients that they’re getting the work they’re being billed for. Capgemini N2K is the users’ front end.
In the first year after beginning the GFS rollout, the system ran smoothly. But as more countries came online—and particularly when it was time to close out big financial reporting periods—performance issues started creeping in. The data warehouse should have been refreshed every 2 hours, but reports were sometimes taking 3 or 4 hours to generate, which meant that the entire reporting cycle could take up to 10 hours. With only 10 countries online but another 30 to come, it was obvious to the IT leadership that they faced a crisis if they did not act.Capgemini Stats
|The number of Capgemini employees globally|
|The number of countries that rely on Capgemini’s Global Financial System|
|The number of projects tracked in Capgemini’s Global Financial System at any one time|
Indeed, system performance had the potential to threaten the adoption of GFS. Many of the affected financial centers had been operating semiautonomously for years, so change management was a critical issue. The delays could become more than an irritation for everyone involved.
“Many of our finance controllers and operations managers had been there for 15 or 20 years, so they liked the old decentralized model,” Jahn says. “When we went to a global, more centralized model, we knew they would not forgive any mistake or regression.”
Varadarajan Subramanian, senior manager, finance, at Capgemini India, is one of those managers and recalls the stress of moving to a shared services center. “We had challenges running reports and getting data out—especially during the peak periods of financial closing,” says Subramanian. “During critical closing periods, performance issues impacted our reporting lead times, and we would have to work around the system with the help of our support team to mitigate wherever possible. But where it was not possible to work around, we had to raise the alarm that we may not meet our deadlines.”
Subramanian’s division was just one of many regional finance departments that ran regular, system-intensive reports against the system. Additionally, ad hoc queries from users around the world were running on a daily basis, making global financial reporting a complex and dynamic process. So while the delays were worse during end-of-year reporting—taking up to 20 hours—they were also causing significant difficulties for users every single day.
“We weren’t delivering performance, and our users’ lives were miserable,” says Elizabeth Rabet, vice president for global finance systems at Capgemini. “Something dramatic needed to happen.”Enhanced Architecture
In 2011, the company’s IT leaders launched a proof of concept to determine the best solution to the performance problems that were causing reporting delays. The test interrogated the speed of Oracle Exadata Database Machine during a vendor review. According to Rabet, there was no comparison. “We saw that with Oracle Exadata, we were reducing the turnaround time of the reports anywhere from a factor of two to eight times,” Rabet says. “Other technologies did not come close.”
The value is in freeing up resources to do additional projects, continue to meet business objectives, and move forward.”–Elizabeth Rabet, Vice President for Global Finance Systems, Capgemini
Oracle Exadata Database Machine is a data warehouse platform that can run queries at up to 10 times the speed of other platforms. It helps centralize and extract mass data on a daily basis from the Oracle E-Business Suite–based accounting and financial system to accelerate reporting of more than 100,000 customer projects each year.
“The major advantage of the Oracle Exadata machine, architecturally speaking, is the cache,” Rabet says. “The storage is in the machine itself, and the networking is faster, too. It’s just a totally optimized machine.”
The Oracle Exadata machine went into production in 2012 and improved turnaround times fourfold. Two years later, the GFS has been rolled out to 80 percent of the company’s locations—up from 30 percent initially—with report volumes increasing by 100 percent. Despite this expansion, the Oracle Exadata solution has improved user experience and productivity by improving report execution with real-time database tracking and detection of the most bandwidth-consuming queries.Foundation for Change
Subramanian says that the regions have seen the impact of the improved performance in their daily roles, reducing the help they have historically needed from global shared services to get through the busy times. “There has been a marked improvement in our ability to run these reports during the closing period over the past of couple years, even as we added 10,000 new employees to our ledgers,” he says. “Support’s involvement has been reduced to a bare minimum.”
Rabet says her organization benefits from staff’s newly available work hours. That new efficiency helps further reduce the total cost of ownership of the Oracle Exadata system.
“Before we moved to Oracle Exadata, the data warehouse and N2K system required constant care. I had to have three people, full time, just babysitting,” Rabet says. “As soon as we moved to Oracle Exadata, this was no longer the case. I didn’t require anybody to do anything. The value is in freeing up resources to do additional projects, continue to meet business objectives, and move forward.”
Users like Subramanian tell Jahn that they are now far more supportive of the GFS. “The feedback was just great. It was, ‘Why didn’t you do this before?’” Jahn says. “It has been a real improvement, even compared to what they had locally before the GFS. We can run very heavy reports, extracting a lot of data, analyzing it, or injecting reports into dashboards with drill-down—something that would probably have been impossible without Oracle Exadata.”
But perhaps most important, the project may have given shared services a good name throughout Capgemini globally. Rather than shared services being seen as a restrictive system that inhibits agility at the local level, the success of the GFS on Oracle Exadata shows that shared services can enable best practices and new ideas.
“It shows people that shared services are important, but that we’re not there to restrict them being creative in how they analyze data,” Jahn says. “They may have good ideas, and shared services can help them act on those ideas.”
Photography by Shutterstock