How To Manage A Globally Staffed Project
By Larry Dignan
Managing the development of new software and deploying systems gets more complicated the more spread out the resources.
Managing the development of new software and deploying systems gets more complicated the more spread out the resources. If your applications are developed in Bangalore; your data centers are in Hong Kong and Los Angeles; and your network operations are based in Rio de Janeiro, you've got a management task ahead of you.
Here's a checklist of what you can do, from picking where to best procure the resource, to managing the collaboration on the rollout:
Define the project: Only you should determine what moves to another country. For some companies, it may be a call center. Others may move applications maintenance and management.
Document processes: "Any service or process that's heavily documented and scripted can be moved offshore," says SevenSpace CEO Peter Weber. The catch is that many companies don't have enough documentation so a person in another locale can follow processes easily, he says. Good documentation ensures teams in the U.S. and elsewhere can collaborate.
Determine the processes involved: Being global will require more than a spreading of technology. Gurvendra Suri, CEO of Dallas-based Optimal Solutions Integration, a company that integrates supply-chain software with SAP, says companies need teams that can process visas and navigate international legal and accounting conventions. Each country is different and has its own quirks, says Suri, who is expanding a facility in Bangalore and building one in Munich.
Define the talent needed to program and deploy. Do you need a coder right out of school, or someone with more experience? According to Forrester analyst Stephanie Moore, an Indian programmer out of college will run about $5,800. Prices escalate from there for programmers with experience.
Suri says his engineers in India earn about $30,000 a year and he could only hire half of them locally. He required engineers with 8 to 12 years of experience in integrating supply chain and enterprise planning software. He hired workers in the United States, Britain and Singapore who wanted to move to India.
Define and procure network hardware infrastructure. Suri says hardware shouldn't be overlooked when moving offshore. A virtual private network router for his India facility ran him $2,200 compared with the $800 it would have cost him in the U.S. Why? He needed hardware quickly and didn't have the appropriate licenses so he had to buy equipment from a local reseller, which could raise prices due to lack of competition. Now he has the licenses, which took weeks to land, Suri can ship less expensive equipment from the U.S.
Compare quality and cost of country. When it comes to quality, offshore workers should have ISO 9000/9001 certifications, which imply an engineer knows international standards for measuring quality in design and development projects. Another certification to look for: Expertise in the Software Engineering Institute's Capability Maturity Model (CMM), a process designed to improve the rollout, management and optimization of software technologies. Indian offshore firms tout their workers' certifications.
Companies that want to cut offshore risks may consider, "near-shore" alternatives such as Canada, where you can cut costs 15% to 25%. Electronic Data Systems uses operations in Brazil to be in the same time zone. Other companies pick South America for offshore outposts to be near existing subsidiaries.
Evaluate obstacles such as communication, airport infrastructure, power and politics. Politics, infrastructure, management hurdles, language and the labor pool all play a role in managing projects globally. Suri says power supplies are increasingly important when moving data centers to India.
China is cheaper than India and other locales, but Moore says companies shouldn't consider it as a viable offshore alternative for software development and maintenance support. She cites the lack of English language skills, legal and regulatory environment and lack of intellectual property laws as risks.
Develop a coordination plan. Suri says it's mandatory to have traffic cops in the U.S. and abroad to keep developers on the same page. He advocates having a systems architect and project manager onsite in the U.S.with peers offshore to keep documentation current and to communicate to development teams. Tools such as Lotus Notes and instant messaging should be used to collaborate and track changes.
Plan every step of the deployment. Ideally, a project will hit the goal line with collaboration by U.S. and offshore teams working around the clock at the installation site. If that's the case, costs can be cut on tasks such as systems integration and testing. If a project is developed totally offshore, the deployment is more complicated. It has to be tested on the onsite systems and it'll take a team onsite longer to deploy and work out kinks.
Monitor the project. The onsite liaison's main job is to monitor the project, keep it on track and coordinate with his offshore counterpart. In weekly status reports, metrics following milestones, documentation changes, architecture tweaks and timelines should be delivered. Optimal puts these metrics in a dashboard.
Manage, manage, manage. Once the project hits maintenance mode, a liaison should stay behind since changes will always be required. Goals will be set by service level agreements tracking uptime, repairs and restoration of service in case of an outage. "To make these projects successful, there always needs to be a U.S. presence on site," says Suri.