by Ben Worthen

Bold IT on a Budget

Aug 15, 20058 mins

Not every organization can afford to spend millions on cutting-edge technology. The challenge for these companies is to find a way to be bold without breaking the bank. Methods employed by our CIO Bold 100 honorees range from the practical (such as limiting the scope of major projects) to the more risky (including relying on beta versions of software) to the outright counterintuitive (developing prototypes with no budget). Each strategy comes with a set of trade-offs, but for our bold CIOs, managing the risks was better than spending the extra dollar.

Fulton County, Ga.

Goal: Make the county one of the most technologically advanced in the country.

Risk: Initiative relies on emerging vendors and beta versions of software.

Mitigation strategy: Testing, testing and more testing.

When Robert Taylor was named CIO of Fulton County, the IT department was so behind the times that the county shelled out money for 1,200 MindSpring accounts rather than support a corporate e-mail system. “My charge was to drag the county kicking and screaming into the 1990s,” says Taylor.

And it was already August 2000.

Since then, Taylor has transformed Fulton County, which includes the city of Atlanta, into one of America’s most technologically sophisticated local government organizations despite an annual budget of just $24 million. The key, he says, is building relationships with vendors, working hard to find price points and payment plans that work for both parties, and making up what you can’t pay for with sweat equity.

For instance, instead of spending $300,000 on a security product from Cisco, Taylor agreed to beta test a new Microsoft product. In another instance, he agreed to help a startup called CloudShield develop the final stages of its software in exchange for a discounted price. These not-ready-for-prime-time products are always tested in Fulton’s lab, then piloted in the IT department and a few other friendly departments before Taylor rolls them out to all 42 county departments. Because his team always comes up with improvements, Taylor compares the process to buying the frame for a house, but putting up the sheetrock and painting it yourself.

This approach doesn’t work for every potential partner, says Russell Mobley, Fulton’s interim deputy director of IT over applications. “Some vendors say, ’We’re a tier-one company, so we won’t negotiate,’” he says. “That’s fine. There are tier twos and threes that have pretty good mousetraps.” And those companies may be willing.

Taylor says that no cost is too small to cut. When this reporter followed Taylor’s instructions to call him at his office, the CIO joked, “One way we save money is to have the other person initiate all our phone calls.”

Questar Gas

Goal: Replace an antiquated customer information system (CIS).

Risk: In the utility industry, projects such as these often take too long and go over budget.

Mitigation strategy: Assembled a top-notch team that spent a year and a half determining the requirements for the project and negotiating with the vendor.

In 2001, Shahab Saeed, then the new VP of IT at Questar Gas, walked into his CEO’s office and announced that the company’s legacy CIS was a time bomb. It was decades old, written in languages that schools no longer taught and had enough bolt-ons to make Frankenstein’s face look smooth. The CEO was concerned (rightly) that projects like that tend to go seriously over budget. But Saeed convinced him that he could pull it off by putting the right people on the team. And he did, completing it two months early and $10 million below budget.

The key to Questar’s success was meticulous planning. Before making any investments, Saeed and his team spent a year studying past industry CIS failures and used these lessons to develop a set of requirements for Questar. For example, the team found that most utility companies tried to convert too much data from the old system into the new one. So when the business said it wanted to move three years’ worth of historical data, the team pushed back, asking for proof that all that information was necessary. A check of the numbers revealed that customer data older than 13 months was accessed in only 1 percent of inquiries. Compromising, Saeed decided to transfer 13 months’ worth of data.

“Every decision was based on facts,” he says. “If someone said that they wanted something, we would say, Fine, show us why you need it.”

Questar evaluated 10 vendors to find the one whose product best matched its requirements and then spent six months negotiating the contract, making sure it contained one specific clause: Questar would not hire any contractors. The people working on the project would have to work for either Questar or the vendor. “We wanted people who had skin in the game,” says Saeed.

Saeed was promoted to president and CEO of Consonus (a Questar subsidiary) and VP and COO of Questar Energy Services and Questar InfoComm (another company subsidiary) in February 2004. “I told my boss that I would accept only if I could remain the project sponsor,” he says.

CUNA Mutual Group

Goal: Reduce the cost of IT for the Members Financial Services business.

Risk: Replacing a system that was working.

Mitigation strategy: Did not sunset the current CRM system until the custom system was ready for production.

When CUNA Mutual Group Vice President of Business Technology Sean Fallon reviewed the IT budget for the company’s struggling Members Financial Services (MFS) division in early 2003, one line item stood out: A nationally deployed Siebel CRM system accounted for 75 percent of the business’s IT expenses. To a certain extent, that was expected. The business’s focus is to provide financial advice to credit-union customers, the sort of people-focused task for which CRM systems are designed. Still, the business couldn’t maintain the Siebel licensing and maintenance costs, and turn around its financials. So four programmers were appointed to build a CRM system specifically for the MFS business. The new system would require an initial investment of just under $1 million, but over the long run, it would be dramatically cheaper than Siebel’s.

Fallon says that the project worked for two reasons. First, the executive team understood that having spent a lot of money on the Siebel system wasn’t justification for continuing to use it if another option could deliver a better total cost of ownership and ROI. Second, the team limited the scope of the project to the functions that the MFS business needed immediately and didn’t allow the developers to get bogged down trying to write functionalities that may be needed down the road. In this way, the development team avoided such details as how the customer contacted the office or what other services he has with the company. All in all, the homegrown CRM system has half as many fields as the Siebel system, which makes it easier to use and cheaper to maintain.

“Most projects have an elegant vision,” says CUNA Mutual’s CTO Rick Roy. “But execution is where it really happens or not. By doing it incrementally, we managed to turn a leap of faith into something real very quickly.”

ING Insurance Americas

Goal: Develop one data model warehouse for six countries.

Risk: No budget meant stealing time and resources from other projects.

Mitigation strategy: Reused code from other projects, and staff worked overtime—willingly.

One might not expect to see an organization with a $450 million annual IT budget in an article about being budget-conscious. But ING Insurance Americas CIO David Gutierrez has built an organizational culture designed specifically to counteract the entrepreneurial malaise that can come from having a large nest egg to fall back on. When Gutierrez and his team decided to build a data warehouse system that could integrate with the legacy back-end systems in all six countries where ING Insurance Americas operates (Canada, the United States, Mexico, Brazil, Peru, Chile), he challenged his IT team to develop the prototype architecture—a $2 million project—without any dedicated resources. It was either that or wait a year to try to get the funding in his next budget.

One of the first questions that people in IT ask when they are assigned to a project is, What resources do I have? says Gutierrez, “but not everything requires resources. It’s like the inventor who stays up all night working in his garage. If you have a deep belief in something, you work on it. If you are too facile with resources, you lose some of that entrepreneurial spirit.”

Programmers were able to develop the architecture that eventually became the foundation of the data warehouse system by including parts of the development in other projects and then reusing them. The staff also put in overtime to work on the project. Once Gutierrez could show the rest of ING’s executives the finished architecture, he was able to secure funding for the full project. “There are many good ideas waiting in the wings for funding,” Gutierrez says. “We don’t want ideas to go away just because a project isn’t funded, so we encourage people to do it with whatever they have.”

Senior Writer Ben Worthen can be reached via e-mail at