With a career path that's traversed studying protons to pension plans, Gian Di Loreto, president of Loreto Services & Technologies, based in Oak Brook, Ill., may be considered an unlikely benefits consultant.

However, Loreto, a doctorate-level physicist, left his lab coat behind in 2000 to aid defined benefit plan sponsors in identifying and eliminating data errors and waste in their plans. His client list includes such corporate giants as M&M Mars, Verizon, Sprint, Kaiser Permanente and Continental Airlines. EBN recently spoke with him to learn more about how employers can save millions through pension data mining and cleaning.

EBN: You're an experimental physicist by training and began your career studying proton/antiproton collisions. How did you evolve to your current profession?

Loreto: I went to grad school as the path of least resistance, but became truly interested in science. However, although being an academic has its advantages, the downside is you're never going to make a lot of money.

So after I graduated, I took a "regular job" as a programmer doing financial reporting. From there, I went to Arkidata Corporation, where I worked as an analyst writing code to fix data-quality problems in pension plans. I found the data was not good in these systems, and it was costing companies a lot of money. So I created Loreto in 2004 to do database cleaning for DB systems.

EBN: What exactly does database cleaning for these plans entail?

Loreto: We're generally hired by companies to work with HR/benefits and IT departments to examine all pension plan data and identify problems. With a [large employer], we generally can find about 30,000 participants with "material problems" - paying the wrong people or paying the wrong amount.
Overall, between one-third and one-half of participants maintained by a company have a material deficiency, and typically, it's a systemic problem rather than a data entry problem. So we identify those issues and fix them.

EBN: How does such plan waste and errors occur?

Loreto: Over the course of an employee's tenure, their pension plan goes through between 10 and 15 new systems from acquisitions, mergers and system upgrades. Often, data is not accurately moved from one system to another. In most cases, the old data is lying around somewhere, and we can use it to correct benefit calculations and save companies a lot of money.

EBN: How much do employers stand to save by cleaning their pension plans?

Loreto: American companies are losing billions each year from these errors, let alone hours of HR professionals' time. One of our clients, Sprint, had 40,000 to 50,000 participants, and had budgeted $2 million to fix problems of benefits being over- or underpaid. We fixed the problem for about $400,000. So that alone saved them more than $1 million.

EBN: How do you view the future of your business?

Loreto: Data quality is such a complicated problem, so [to resolve it] employers have to make sure they have a specialized team of people where that's all they do. Awareness is increasing. Five years ago, I had to make the case for this service, but now people are more aware of data quality and how important it is for their plans.

To learn more about Loreto Services & Technologies, visit www.loretotech.com.