Making the Grade

Approaching Northern Trust's Senior Vice President Mike McGlinn with an STP-project proposal can feel a lot like handing in a high-school English essay. In both cases, you're going to get a grade.

That grade, ranging from the customary A-F, gives hopeful project managers a pretty good idea of whether they can expect any funding to come their way. In fact, as evidenced by comments from JPMorgan Chase and Lehman Brothers' executives, it appears many firms in the financial-services industry have taken the scholarly approach to prioritizing STP projects.

"We also have a grading system," said Diane Eshlemman, managing director and chief procurement officer with JPMorgan Chase. "We use it to evaluate investments. First we look for instances where we need to do system upgrades anyway, either because we have technology that needs to be improved or because we have redundant systems."

McGlinn says that Northern's A-level projects are those that must get completed no matter what, while projects that are marked B- receive only the barest life-support-level funding. "If you are a C or below, you might as well not come back because the well has dried up for you," he explains. "So really the higher priority projects we guard, not just with moats but with anything else we can, and new projects that try and get funding away from those are judged pretty harshly."

Lehman Brothers Managing Director of Global Operations and Technology Bob Schwartz says he likes to categorize STP projects into four areas. The first involves infrastructure projects that are "the hardest ones to justify in many cases." He says these are long-term projects such as reference- or static-data platforms where the firm may not see an immediate benefit because so much integration is required.

The second area involves projects that are client-driven. Those, he says, are easier to justify. "If a client says we want to interact with you in this particular way, that comes to the top of the list," says Schwartz.

The third type of project is one with an immediate payback, which could come in the form of headcount savings. Not surprisingly, he says, those receive quick approval. "We have a very expense-oriented culture at Lehman Brothers and whenever we come across a project where the benefit is obvious and quick, we will fast-track it," says Schwartz. "There is almost unlimited funding for those projects."

Finally, the fourth category involves projects that will improve service but have a hard-to-prove ROI. Schwartz calls these "Field of Dreams" projects alluding to the "If you build it, they will come," theory espoused in the movie. If these project don't get the green light, however, he says there are other ways to convince purse-string holders to make and investment.

"The last resort is to point out to management that we are not six months behind the competition and very soon what we want to build the client is going to be asking for," adds Schwartz, "so that may push those kinds of projects over the goal line."

Bernard Donefer, associate professor at Baruch College's MBA program in New York, says reaching the goal line starts with laying out an overarching vision for STP. He explains that many ancillary STP projects with difficult-to-calculate ROIs need to be justified in the context of that vision.

"Those projects that don't have immediate payback fall into this category. So you can explain that we need to do this because it is the glue that will hold together the entire project," says Donefer. "I think trying to build STP on a project-by-project basis is very difficult."

Eshlemman agrees, noting an overarching STP vision must also bridge the silos of front-, middle- and back-office technology that segment most financial-services institutions. "In the past, they have been allowed to build separate applications, separate data repositories with separate protocols as they worked in their towers," she says. "But unless you have an enterprise-wide view of STP you will continue to have investments made locally that are not leveraged on an enterprise basis."