E-Trade's CTO talks about what Linux means to his company.

InformationWeek Staff, Contributor

November 20, 2003

10 Min Read

WS&T: What did you spend your technology budget on this year?

LEVINE: We spent all of 2002, and the first half of 2003, looking at consolidation opportunities - not contraction, but more how can we do more for less, and not as a corporate mandate, but as a technology-enablement initiative. What we saw was that the technology deflation continued to occur in hardware and software, and we wanted to really tap into that deflation, which was happening mostly in the Intel world at the time. What we spent money on in 2002 was trying to standardize our architecture and move to more generic architecture. So we moved more into Intel and Linux and away from proprietary operating systems, which cut our maintenance costs.

In 2003, our biggest initiatives have been to rebuild the back office for our brokerage system. We have a new banking system and we've worked on a new CRM initiative. We're an early adopter of CRM and trying to help our customers. We've now morphed that into a more generic data warehouse for our customers' (information), to work on things like campaign management, to try to cut down on e-mail and mail.

In addition, we've got our active-trading product. We have a large active community, and we want to keep a lot of new products in their hands because they demand it.

WS&T: E*Trade has gotten much attention for being a leader with Linux. How is your firm reaping its benefits?

LEVINE: Linux is kind of a metaphor for the kind of computing that is a revolution right now. You have 20 years of the most incredible research and development in Intel architecture. It's probably the most heavily invested architecture ever. The result has been an equally tremendous deflation surrounding that. We realized that we had to use an Intel architecture to really gain the power of that deflation, and get on board with cheaper storage and cheaper computing. We asked (ourselves), what would it take to port to Windows? We have a Unix-based environment, so it would be easy to port to Linux. We ended up building a pilot of the system. Linux really is a metaphor for tapping into lower-cost computing and making computing a real commodity so we could focus our capital on what is important. You have a lot of flexibility when price of computing drops tremendously. There's so much power in that kind of low-cost platform. One thing that is beautiful about Linux is you are no longer managed by the vendors. You manage the vendors! If you don't like one vendor, you can move to another. Most shops are developing on Linux and porting to all other environments.

WS&T: What about on performance? Is Linux scalable enough? Linux scalable?

LEVINE: We haven't found any issues. There's a new release coming out that gets rid of scalability issues and its only getting better. If you have a lot of custom-development " people are struggling with that legacy aspect.

WS&T: What is your project-prioritization process?

LEVINE: We have a process that the businesses and technology walk through from inception all the way through to project initiation, definition, specification. It sounds rigid, but it's a nice process of businesses talking about what they'd like to do and technologists working as partners, figuring out how that translates into projects and costs. Then we sit on a weekly technology-investment committee at the firm where we go through those projects and make decisions about whether we want to invest our capital that way. It's an excellent process and it helps us really get the businesses to focus on what's important for them. They're the ones that are driving how we do things and it helps us keep track of what are we doing -- when should we start to add resources, or contract on resources. Everything is articulated and discussed. It keeps dialogue between all of the stakeholders. Whatever the process is, it must be about communication.

WS&T: How do you determine return on investment?

LEVINE: Return on investment is probably the most important way to look at (investments). The business has its own the initiation process. They articulate what it is they want to do and they assign a return on it. So before it even gets into the IT world, it goes through business management. Through a process of articulation, it gets them to see technology as a driver for revenue enhancement and expense reduction. Revenue enhancement is easiest project to see ROI. Expense reduction is also pretty easy to measure. Regulatory often has fines associated. You know that if it's risk avoidance, you have to do it, so you figure out how much money would be at risk if you didn't.

WS&T: Where are the largest costs in taking on a project? LEVINE: I think the largest money cost is certainly on the human resources.

WS&T: What is your perspective on outsourcing versus staffing a project in house?

LEVINE: We really have a very small consulting force, probably under 10 percent of our overall staffing, and only have 60 contractors out of 775 people. We find that to be a motivating factor. Our employees are motivated to be customer focused.

We keep looking at where we can really be successful (in outsourcing). From our experience, if we can isolate an overall project or segment then we'll gladly outsource, like our help desk. We have some technology projects that we are in the process of outsourcing that are fully encapsulated and easily documented so easy to pick up on. We have products that we have a team of people working on that are not intertwined with rest of our environment. If you can take a whole project with managers, programmers, and analysts and put them altogether in the same environment then that's a real win.

WS&T: What about outsourcing infrastructure engineering?

LEVINE: We're just not there yet.

WS&T: Is there an element of innovation that an outsourcer can contribute that an in-house employee might not have thought of?

LEVINE: I think on paper that's true, but the number one issue that technology faces in terms of delivering what they need to be delivering is the connection to the end customer. You lose that with an infrastructure professional that doesn't know about your customer. You lose that connection. You could have the greatest network engineer build the greatest network of all time, but if juice cans and strings are what the customer asks for, they're going to miss that.

WS&T: How do you evaluate third-party solutions when going to market for an application?

LEVINE: First someone articulates that need. People sit down and look at different vendors and it comes right back to investment committee, which meets once a week. It's a heck of a lot of fun. People are talking about projects and our production calendar of when we're going to roll out new products and changes. Two thirds are technology people (senior architects and engineers), and technology people love to discuss different technologies.

We have an integrated technology plan so we don't have product silos with disparate technologies running in the closet down the hall somewhere. When you look at etrade.com, you don't know what is behind it, and where it is, and how it all works, you just expect it to work. All of our products come together at the customer. We work very hard at trying to make our customer the central focus of the technology, not our organizational chart. All of the projects are product-aligned in development. You have incredible product knowledge with technology people who are aligned with their product.

WS&T: What are some emerging technologies that might get some play time in your firm in the next year?

LEVINE: As a financial-services company, what we've been really looking at is our brokerage customers. They want an electric experience, everything alive: news that scrolls, and is fresh that you don't have to refresh, market data that updates, alerts and notifications of all different kinds - just like the institutional brokers and traders have. The retail investor is interested in that as well. We're thinking about ways of giving this to more and more of our customer base. You'll see a lot of those kinds of initiatives from E*Trade in the next year.

It's an exciting year for new features and functions that show the promises of having an Internet-based delivery service for financial services. 2004 will be the year that everything starts coming together and people really start to see the power of having information at their fingertips electronically.

WS&T: How much emphasis is placed on customer relationship management?

LEVINE: One thing that we've started to see real success on is campaign management, communications to customers in a closed loop to ensure that we're not over-communicating. On top of that , there's a lot of new initiatives that are built on what customers wish for. We do a lot of customer segmentation, providing great service for customers that really need that, or one-to-one contact for customers that really require it.

WS&T: What major IT projects are on the table for next year?

LEVINE: We're early in the cycle of next year so everything looks like a wish list. We're consolidating our back office. We announced that we're moving to ADP on the brokerage side, which we expect to do by this spring. We're also looking at our banking back office, and looking at various alternatives for banking back-office systems. Risk management and risk-management systems will be a big initiative. We just announced currency and futures trading so there is integration of those for our active traders. We're also spreading out to our less-active traders so that (normal investors) can get access to more hedging products.

Also this year we will be looking at our middleware and database technology. We currently use TUXEDO for our middleware for all of our E*trade products. We also use Sybase and Oracle on the database side, so we're looking at those technologies and how they fit in with our overall architecture. And we have to spend less money as well. We're going to try to keep doing more for less so that we can remain competitive and give great pricing for our customers.

WS&T: what is your biggest challenges are for 2004?

LEVINE: Our biggest challenge now is trying to keep up with all that we want to do. I think that we really want to get our infrastructure set for the next five years of development, and get our back offices set, so that we can offer this integrated, cross-sold product for our customers, and make it seamless for them.

The challenge is to keep that focus going and figure out the things that we can do. It's not priority setting, but managing our own expectations of what can be done in one year.

Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.

You May Also Like


More Insights