Page 4

By Peter Henig

The Efficient Frontier: AXA Financial Inc.

"When I got here, the whole model of asking for money was based upon what I referred to as 'emotional appeal.' Whoever screamed the loudest usually ended up with the most," says Paul Bateman, 50, and in his fourth year as director of enterprise governance for AXA Financial Inc. AXA hired Bateman from Citigroup in April 2000 to tackle its IT free-for-all. At AXA, the $7.6-billion U.S. subsidiary of AXA Group, the Paris-based financial-services conglomerate, IT had long been seen as a competitive differentiator—and remains so—but at $300 million, the company's 2000 IT budget would soon equal 70 percent of its total net income.

Like so many businesses, AXA had a great fin de siecle. Entering 2000, AXA's revenues soared as the financial markets boomed, and the company took in $2.4 billion in cash (and $5.6 billion in stock) for the sale, to Credit Suisse Group, of its 71 percent stake in Donaldson Lufkin & Jenrette Inc. Yet by the end of 2001, things had changed dramatically.

Though AXA Financial's revenue reached $7.8 billion for 2001—more than $1 billion ahead of the previous year—earnings fell off a cliff. With the economy contracting, but especially in the aftermath of Sept. 11, reported net income declined more than 80 percent, from $2.4 billion in 2000 to $425 million in 2001. AXA's insurance firms not only faced claims against losses in the terrorist attacks, but AXA Group itself owned a 25 percent share of American Airlines that morning.

Early in 2002, Henri de Castries, AXA Group's chairman and CEO, reassured shareholders that AXA would continue to pursue its overarching strategy of combining its global scale, management expertise and extensive IT to sell mutual funds, say, to its annuities customers. But, internally, the status quo wouldn't do. The message was out: The firm needed to do more with less. And efforts launched in the spring of 2000 to bring greater order to AXA Financial's array of IT projects took on renewed urgency.

"When I took an as-is snapshot of what the lay of the land looked like," Bateman recalls, "I discovered there were a number of projects behind schedule and over budget and, more important, that all projects were created equal. There was no hierarchy of improvement. There was no sense of how these projects related to our vision, mission, long-term, short-term—any kind of objective you wanted to define."

To better understand where AXA spent on IT, and to link its budgets to strategic, rather than emotional goals, Bateman bet on a more empirical, logical approach to IT oversight—the much-hyped, oft-misunderstood technique of portfolio management.

"What appealed to us about portfolio management was that we needed a disciplined approach to our IT projects," recalls Charlie Curcio, AXA Financial's CFO for IT. "If you have spending on the rise, such as we had in the late nineties, when e-business was booming, you need a new way to focus on the true value of all of your projects."

Some of the initial stages of IT management—taking an inventory of existing projects, and assessing budget and on-time completion—didn't require more than spreadsheets. But as the cost-budget analysis of pending projects got more complex, Bateman turned to New York City-based software company UMT (United Management Technologies), whose approach to portfolio management has its roots in the theories of Nobel Prize-winning economist Harry Markowitz. Whereas some portfolio-management programs are little more than project management on steroids, UMT supports managing IT in much the same way that a fund manager does an investment portfolio. As such, it offered Bateman a way to decide not only how to optimize the performance of existing projects, but which projects should be approved in the first place. And it allowed him to evaluate any area within IT where money is managed and returns on that money can be optimized—particularly when he had to choose between a range of alternatives and make trade-offs. If I invest in X, for instance, what will be the impact of not investing in Y—and will I or won't I be best able to execute on strategy Z?

"As the market continued to contract," Curcio says, "portfolio management became increasingly valuable, because of the emphasis it places on getting the most out of the same resources."

If Curcio's endorsement sounds familiar, that's because portfolio management is enjoying something of a vogue. "Consumer goods and automobile companies are starting to use portfolio management," says Doug Lynn, a vice president at META Group Inc. "Big pharma has done it intuitively, but they haven't used it to the extent they will in the future. And aerospace and defense are very big on this because they need to make sure projects are finished on time and on budget." Lynn's firm is quite bullish on portfolio management, predicting that the market will likely double, to $350 million, this year. Moreover, the benefits in terms of cost savings and efficiencies are reported to be remarkable.

"Leaders in IT portfolio management have demonstrated ways to cut IT costs by 2 to 5 percent, improve productivity 25 to 50 percent, and shift 10 to 15 percent of their IT budget into more strategic projects," claims Eric Austvold, a research director at AMR Research Inc.

To download a chart of AXA Financial's Portfolio Management Process, click here.

And yet, when Bateman, Curcio and a few others at AXA selected UMT's approach, it wasn't such an easy sell. Sure, it helped that people at AXA understood investment portfolios, asset allocation, risk management and the like, but not everyone saw its appeal right away. Swift results, however, helped Bateman gain traction.

"I bet we saved $5 million to $10 million in the first year alone," Bateman says, "on projects that would have automatically gone through before. But now the business units knew what could get killed, so they killed it first. It just became immediately apparent how much junk we weeded out of the system."

Page 2

Getting the Buy-In

In France, AXA Group has brand recognition similar to IBM Corp. or Citigroup. AXA Financial, however, is somewhat less well-known in the U.S. Coming in to work on Sixth Avenue in New York City, Bateman didn't have a clear sense of what the culture of the place would be like, or if his new colleagues would be receptive to the changes he'd been hired to execute. He's in charge of provisioning IT for several multibillion dollar companies at once, and while some of them didn't have to be convinced of the need for change, others, the cash cows, wondered why Bateman wanted to repair what wasn't broken. Right away, it became obvious that without then-chairman and CEO Ed Miller's buy-in, no one would have paid him any heed. "If I didn't have the CEO's endorsement, forget it," he says. "I wouldn't even have been able to compete with Sisyphus."

Fortunately, Bateman got the endorsement, and if Miller got the ball rolling, Christopher "Kip" Condron, who replaced Miller in May 2001, has been the one to take portfolio management to the next level. Condron formed, and chairs, a corporate oversight committee composed of executive vice presidents and C-level executives, and the committee uses UMT's software.

"IT portfolio management is not just another market category that vendors or analysts dreamed up to sell more software," says AMR's Austvold. "In fact, it is part software, part organizational discipline. It is an evolutionary improvement for all project management software, adding capabilities to judge an entire collection or portfolio of projects on their relative health, importance, return and risk."

Before any evolutionary leaps could be made at AXA, Bateman first gathered the best intelligence he could get from each of AXA's six U.S. business units. Bateman describes how the process used to work: "At first, the vice president of the field group would come in and ask for one thing, then the product guy would come in and demand another thing. So whoever had the most passion behind their pleas generally got the higher levels within the budget." Not only were executive vice presidents asking for different things, but IT essentially had no clue what the various business units were trying to achieve. For example, if a sales unit was moving toward a model requiring more customer self-service, could there be a cheaper, more efficient, off-the-shelf investment, as opposed to one developed internally? IT had never asked, and sales had never volunteered its business intentions.

So the first step of Bateman's approach to IT management was to take an inventory of each business unit's IT needs as they related to its individual business objectives. At the same time, Bateman asked Condron to clarify each division's strategic objectives. Then the corporate-level objectives had to be defined and identified in a empirical way—specific targets had to be set so that the goals could be plugged into AXA's new prioritization, valuation and optimization model for IT investments. Each executive vice president, and his or her respective group, was asked to determine the level of impact each project might have—extreme, strong, moderate or low. Would a proposed IT project, for example, have a strong or moderate impact on the overall goal of reducing costs 10 percent? Increasing sales by 20 percent? Improving customer satisfaction by 5 percent? Does upper management share all of those objectives? If not, does management value one project more than the others?

Page 3

Interestingly, what first appeared wasn't just one list of projects and weightings for each business group, but two. One was the executive vice president's own list and assessment of the impact of each project; the second list was one compiled by each executive vice president's direct reports. At each step in the process—taking inventory of proposed projects, evaluating them against alternative projects, measuring their impact against business objectives, and then prioritizing the results as aligned against business objectives—UMT's software provided the basic platform and pre-programmed algorithms for Bateman and company to plug all of their data into one comprehensive model. Once the inventory, business objectives and impact conclusions were entered into UMT's applications, Bateman was able to show the head of each group a "strategic alignment report" for each list. If the two reports didn't match up, the vice president had more than an IT or budgeting problem on his hands. "If he sees his alignment results are far different than those of his [business unit], he's got a communication problem as well," Bateman says. "Either he didn't explain what he wanted, or they didn't get it."

Running the lists through the system again, Bateman generated a report to determine which projects had the highest priority for budget approval. At the same time, IT would give rough estimates on what it would cost to develop each project. With the prioritization reports and the IT estimates, UMT's software model could then apply budgetary constraints against the prioritization report. For example, what happens when there's only $40 million in the budget for a product group, when the budget was $80 million? Having loaded AXA's financial constraints into the model, out would come a list of all the projects, showing the highest potential for return in terms of ROI and business value based on the $40 million limit.

Based on his analytics, Bateman supported a new life insurance product but curtailed a customer self-service offering. "At the time," he says, "a high emphasis was placed on new products versus new service programs."

Though UMT's automated rules for prioritizing might seem rigid, the portfolio-optimizer software is flexible enough to force a project back into the system that it may have initially declined, but only if the executive vice president can argue that such a project is of greater short-term value, or fulfills a clear regulatory requirement.

Ultimately, Bateman says, the most difficult part of the process wasn't getting AXA, as an organization, to understand what he was trying to do, but rather getting everyone to accept it. "The difference was challenging a mentality and culture that was used to saying, 'Oh, our competitors are doing it, therefore let's do it,' without worrying about the financial repercussions or payback."

Page 4

Keep Evolving

Three years later, AXA now appears ready to make the next evolutionary leap, moving forward to the highest level of what UMT CEO Mike Gruia calls the "portfolio management maturity model." Bateman's first move had been to install a system of governance structures; his second was to convince the organization that portfolio prioritization not only works, but is non-negotiable if the company wants to achieve its stated business objectives.

Now, he's in a position to apply the model not only to IT, but across the organization—asking everyone, in effect, "What are you capable of?"

"All sorts of trade-offs and alternatives can be plugged into the model based on different sets of strategic objectives," says UMT's Gruia. In other words, just as a broker might consider the risks of one set of investments as compared with another, Condron and upper management can ask Bateman to determine the best set of IT investments if they want to pursue a strategy of top-line growth, or a more conservative, bottom-line approach, or one emphasizing cash flow.

Management can obtain a different prioritization report for each set of strategic objectives. As the company changes, the model can change with it, creating an ever-expanding set of scenarios based upon the growth, stability or retraction of the company.

In terms of cost savings, Bateman doesn't have an exact figure. In addition to the $5 million to $10 million he figures AXA saved the first year, he cites several intangible gains—far better levels of communication, information flow and alignment of resources with overall business objectives. "There are ways that [portfolio management] has helped that I couldn't even put a dollar amount on, because in a way it's actually changed the culture of the way we operate," Bateman says.

What's next? Beyond rolling out the latest version of UMT's dashboard, Bateman envisions a centralized repository of all data as it relates to an IT portfolio, a company's investments and resources, where a CIO could go to his or her desktop and centrally manage everything from within that one model. Of course, that would be the brass ring for bridging corporate governance and IT.

"It gives you the ability to look at any aspect of your portfolio, whether it's people, whether it's timing and launch, whether it's actuals versus plan or approved rates, or return on investment," Bateman says. "Everything from end to end, the holistic process viewed through one microscope."

Portfolio management can't address all the challenges that AXA Financial and its parent, AXA Group, face. Though its proposed acquisition of New York City-based MONY Group Inc. gained approval in May, AXA's share prices have slipped slightly, from a 52-week high of $24 in January to $20 as of May 26. Securities analysts say AXA is benefiting from higher insurance prices and rebounding equity markets, but they complain that the company hasn't reported its financial results clearly, and analysts worry about the firm's estimated $1.1 billion in asbestos and environmental liability.

Many of these concerns are well beyond Bateman's immediate control; portfolio management, he says, has made AXA more effective at what it can control—making the company more efficient. In fact, Bateman confesses surprise that IT portfolio management has only recently caught on. "I'll be honest with you, I don't know why it's taken so long. Maybe it's because people are so caught up in changing platforms and consolidating servers and building the latest operating system," Bateman says. "To think there are companies not even in the infancy stage of this is mind-boggling."

Peter D. Henig is a contributing writer at Option- etics.com. He's also written for Forbes ASAP, Venture Capital Journal, Red Herring, The Economist Intelligence Unit and The Daily Deal.




IT Governance: How Top Performers Manage IT Decision Rights for Superior Results
By Peter Weill and Jeanne W. Ross
Harvard Business School Press, May 2004

White Papers

"Avoiding Investment and Project Failures"
By Val Sribar and Al Passori, META Group Inc., April 2004

"What It Takes To Be a Leader In IT Portfolio Management"
By Eric Austvold and Alison Bacon, AMR Research Inc., July 2003

"IT Investment Management: Portfolio Management Lessons Learned"
META Group Inc., December 2002

This article was originally published on 06-01-2004