Interview: MIT Research Director Peter Weill on IT Infrastructure

By CIOinsight

Interview: MIT Research Director Peter Weill on IT Infrastructure


Peter Weill Interview

The short and painful history of e-business has taught us that execution and capability count—and that's made the issue of IT infrastructure a timely one. Once viewed as mere plumbing, the pipes and pumps that keep data flowing, executives are finally grasping the connection between IT infrastructure and corporate strategy. But what is that connection, and how do you leverage it? That is the question Peter Weill, the new director of the Center for Information Systems Research at MIT's Sloan School of Management, has been working on.

Weill, an Australian, is no wild-and-wooly Web guru or ivy-covered obscurantist. Says Peter G.W. Keen, a prolific author of books on e-business strategy: "Weill's an academic realist. What he says is reasonable, shrewd and realistic, and very well grounded."

Weill initially laid out the strategy and infrastructure link in the 1998 book he wrote with Marianne Broadbent, Leveraging the New Infrastructure: How Market Leaders Capitalize on Information Technology (Harvard Business School Press, 1998). His new book, Place to Space: Migrating to eBusiness Models (Harvard Business School Press, 2001), co-authored with Michael R. Vitale, extends that connection to e-business. Weill says e-business has led to the centralization of IT infrastructure—which is the shared technical foundations on which individual applications are built and maintained—primarily because e-business models require providing a single point of contact across multiple business lines.

The result of a detailed study of 50 traditional companies' e-business initiatives, the book argues that there are eight "atomic" models for doing business electronically—including the "direct-to-customer" model made famous by Dell Computer Corp. and the "shared infrastructure" model exemplified by Covisint LLC, the online auto parts exchange co-owned by the big-three automakers. Depending on their capabilities, companies can use these "atomic" models as building blocks, snapping them together to further—or even create—a business strategy. How does a company's infrastructure and the underlying architecture affect these models? Can infrastructure determine e-business strategy rather than the other way around? Weill explored these issues with CIO Insight executive editor Allan E. Alter.

CIO INSIGHT: How can CIOs bring the most value to the discussion on e-business strategy?

Peter Weill: I think good CIOs are schizophrenic. They have to be very detail-oriented, very good at the implementation side. But that's just the cost of admission. The really good CIOs also offer a way to affect strategy through technology. We've typically said that business strategy drives IT strategy, but e-business has underscored the notion that technology can create opportunities to affect strategy. Really good CIOs have to get value from their infrastructures and integrate new technologies in a way that is business oriented. The most common complaint I hear from CEOs about their CIOs has to do not with their management of the IT dollar, but with their ability to help the business folks use technology better, to strategize creatively.

What questions must CIOs be prepared to answer in order to be effective in talking with other executives?

They need to know what strategic options are on the minds of the heads of the business units. That's the starting point. Another thing is understanding what competitors are doing with respect to technology. You often see the CEO turn to the CIO and say, "Tell us about the technology platforms of our competitors. Can we be hurt?" Both the defensive and offensive views of technology are important.

By providing a ubiquitous infrastructure, the Internet has allowed much more cooperative business models than we've seen before. CIOs are responsible for negotiating all the agreements linking their companies' technology with other firms. Still, many of these shared infrastructures will be a disaster because it's incredibly difficult to share. If you think about Covisint, you can imagine the discussions going on at Ford Motor Co., General Motors Corp. and Daimler-Chrysler AG about what data should and should not be shared.

How can a CIO use the eight atomic e-business models to analyze IT strategic opportunities for his or her company?

Look at the eight models [see Weill's book, Place to Space, for details] and ask: "What are the competencies of the organization?" By asking what you are good at, you can see which models are appropriate. CIOs have to keep in the back of their minds a set of capabilities for infrastructure and a set of e-business models; those two sets have multiple links from one to the other. The good CIOs in an e-business world will sort those links out, and then figure out which capabilities they should source internally and which in the marketplace.

But changing or combining e-business models means creating new infrastructure, new skills, more systems and more integration. Isn't that a tremendous burden?

It's not only been a burden on the IT organizations, but a burden on the capital expenditure of the firm. Infrastructure comes in large chunks, and those are huge hits. What this does is raise the importance of architecture, an integrated set of standards, policies and organizational frameworks that guide a company's IT decisions. Architecture used to be a background operation for a person who was very technical; it's now become a key strategic option. Architecture decisions made today will predict the infrastructure capabilities of tomorrow, which will predict the business capabilities of the next day. We're seeing firms struggle with the governance of architectural decisions: How do you govern architecture when it's now fundamental to the way in which we do business? Firms have attempted doing this by going through committees and working groups, but there isn't yet a really clear answer.

Cisco is a great example of a Net-based architecture. They used ERP as a core technology, but added functionality with net-based applications. That enabled their business at both the supply and company ends.

Interview, continued

Interview, continued

Which architecture issues should be discussed with business managers?

Mostly work architectures—the way in which the organization will do work. What are the standard business processes we're going to use, and what are the standard platforms on which we're going to build those processes?

Do you mean computer platforms?

That means computer platforms, data platforms, telecommunications platforms and applications platforms. Once you go with SAP AG, for example, as an ERP, you set an architecture by default. That's good for firms where the use of the technology is not a competitive advantage. But when technology is a source of competitive advantage, you need to be more active in setting architectures as an embodiment of strategy. And you can't rely on an ERP provider to do that.

Firms are struggling with how to get data out of applications and into some centralized place. I would argue that the notion of a data architecture is now something that is firmly part of the business planning process, or should be. Managers shouldn't be setting their data architectures; they should be setting the governance structure so that sensible data architecture decisions are made by the people who have a stake in the issue.

Business managers might find it tough to discuss such issues. How can CIOs effectively discuss infrastructure with them?

The notion of business maxims or principles works well. The business has a responsibility to identify the principles of doing business, and the IT folks can then translate that into principles for managing IT. Once you have them, decisions about IT infrastructure investments are a lot easier.

Imagine that the business maxim for Johnson & Johnson was to manage its global customers through a single point of contact. That would be a business maxim. So what's the IT maxim? We need to have common and shared data in the following categories: customer data, financial data and product data. And the data should only be entered once and used by all. So the business folks set the business maxims, and the business folks and the IT folks together set the IT maxims. And then the IT folks deliver on the infrastructure to make those maxims happen.

Where do you see e-business heading as we move forward?

The future of e-business has to do with the migration across atomic models. Companies are going to capitalize on successes in one model to move to another. For example, we are seeing companies migrate from the direct-to-customer to full-service provider model, or from content provider to virtual community to direct-to-customer. Lonely Planet, the Australian travel guide company, started as a content provider. Since then it's migrated into a direct-to-customer firm, selling books from its Web site; an e-mail, voice mail and long-distance call service for travelers; and a city guide for the Palm. It is also becoming a virtual community.

What emerging technologies or services provide the greatest opportunity for e-business strategy and e-business infrastructure?

I'm a bit of a mobile e-commerce skeptic at the moment. I'm not convinced yet that we're going to do major transactions on m-commerce. It's still pretty difficult to make it work, and the interfaces we have are not very convincing.

I think voice-activated Internet access will be very strong. I'm very optimistic about products like OnStar, the General Motors concierge service, [which links drivers directly to OnStar customer service staff]. They can provide a whole bunch of services through their call centers, and that's a lifetime revenue stream for GM and just a great opportunity to add to their portfolio of businesses. Integrating multiple sources of data at a point where you can take action, like a call center or an ATM, is the future of technology. I'm very excited about that level of customer service.

What should CIOs do to avoid being the cause of an e-business disaster?

We don't have the unseemly haste anymore to get to market fast. We went through this phase of having to be first because we thought that would create value—but that turned out not to be the case. So I think CIOs can push back a little bit in terms of speed versus quality.

There's been a return to rationality in terms of business models. The disciplines of business models and business cases that we've been using for 10 years went out the window for a while. Now CIOs should be a source of reason and rationality about what's doable from a technical point of view. That's important.


's Eight Atomic E-Business Models">

Weill's Eight Atomic E-Business Models

Content Provider
Provides content (information, digital products, and services) via intermediaries.
Example: AOL/Time Warner, Morningstar, Reuters

Direct to Customer
Provides goods or services to the customer, often bypassing traditional channel members.
Dell.com, Gap.com, RealNetworks.com

Full-Service Provider
Provides a full range of services in one domain (e.g., finance, health) offering both its own products and those of others, attempting to own the primary consumer relationship.
GE Supply Company, Prudential Securities

Brings together buyers and sellers by concentrating information.
eBay, Yahoo

Shared Infrastructure
Brings together multiple competitors to cooperate by sharing common IT infrastructure.
Covisint, ABACUS International

Value Net Integrator
Coordinates activities across companies in a value net (physical or virtual supply chain) by gathering, synthesizing and distributing information.
Cisco Systems Inc., Seven-Eleven Japan

Virtual Community
Creates and facilitates an online community of people with a common interest, enabling interaction and service provision.
The Motley Fool, Inc., The WELL, Parent Soup

Provides a companywide single point of contact, consolidating all services provided by a large multiunit organization.
Colonial Limited, State of Victoria (Australia)


This article was originally published on 07-01-2001