IBM's anti-control freak

Linda Sanford is the senior vice president who runs IBM's on-demand transformation business. In English, that means it's her job to advise IBM and its clients about what parts of the business are worth keeping and what should be handed off to a specialist. It's a timely subject: General Motors last week announced it's handing $15 billion worth of business to IBM, Hewlett-Packard, Wipro and others so it doesn't have to run its own computer systems.

In Sanford's view, the increased use of interchangeable commodity components--whether call centers in India or memory chips for PCs--is a trend to be embraced, not rejected. Outsourcing lets a company whittle away nonessential work.

But the handoff to an outside partner isn't easy. "Of the biggest inhibitors, the hardest element is really around culture," Sanford said: It's tough for companies to let others handle essential work they're used to running. But she argues in her new book, "Let Go to Grow," that not relinquishing control makes companies terminally slow to adapt.

A metaphor for Sanford's approach comes from her own past. When she took over IBM's mainframe business in the dark days of the 1990s, the high-end server line was widely assumed to be headed for extinction. Among the changes she implemented was to support standards from the rest of the computing world rather than being an isolated island of IBM-controlled technology. Some measure of mainframe relevance was restored.

Sanford discussed her ideas with CNET News.com's Stephen Shankland.

Q: To start, could you distill the essence of the book?
Sanford: The way I would describe it is that there have been certain trends here--the Internet, globalization, deregulation--that have fueled a lot of competition and therefore are driving a lot more commoditization in businesses and industries around the world.

As Thomas Friedman says, the world is certainly flat because of that, and so my point of view here is that in order to compete and grow in this world, you need to let go of the old command-and-control paradigm of running businesses and doing it all yourself, because it will only slow you down, and today's world is ever changing on us in very unpredictable ways. So the answer to the world is flat is let go to grow.

So the answer to the new competitive environment that you describe is what you call "value webs." What is it and how does it differ from the old-school "value chain"?
Sanford: Value webs are partnerships driven by what I would call economies of expertise. So you're looking for partners who happen to have expertise in a particular piece of a process that you might normally do, and you might say, "Gee, why am I doing it? Others have better expertise and economies of scale than I do," and so you form a partnership. The company who is forming the partnership wins, the new partner wins and the customer ends up winning at the end of the day.

Why can't you do that under a value chain?
Sanford: A value chain tends to be more within a business. Your value chain is made up of some of your suppliers and your customers and your employees. Value webs really are dealing with whole new businesses, and the expertise that they bring to freeing your resources to focus on your core. For example, you might choose to say, "Why am I doing logistics at all, why don't I go to somebody who has an economy of expertise in this space like a FedEx or UPS and have them do all of my logistics for me?" It's a much bigger piece than just a supplier who would pick up my packages and track them for me as I'm shipping my product.

Where led you down this path?
Sanford: My job in IBM is to reinvent ourselves. I've been spending a lot of time inside the company understanding what we do or how we do it, listening to employees so that we can work on reinventing the things that will make us more agile and more flexible. In a comparable way I've spent a lot of time with our customers in looking for their input to how we can be more responsive to their needs and at the same time learning from them some of the things that they themselves have done in order to drive growth in their particular businesses.

Since this is your bailiwick inside Big Blue, can you give examples of things that IBM has let go of?
Sanford: One of the first things we did, several years ago, was let go of our in-house manufacturing. We found that there were others out there who had an expertise and a scale that we didn't necessarily have that could really drive much more efficiency and effectiveness in that process. We have many partners around the world today that do a lot of our manufacturing for us. (Editors' note: Although partners build some products, such as Intel-based servers, IBM still manufactures its own high-end servers.)

Another example would be in our human resources. For some of the administration of our HR programs, we partner with Fidelity. It frees up our HR professionals to really work on what is core to our value proposition: how to best identify, attract and retain the best talent around the world. And part of our value web is actually within the company in our global services organization for some finance processes, like payroll or accounts receivable and accounts payable. We have our global services colleagues run those processes for us as well for as other customers.

IBM has had financial troubles with microelectronics. I'm guessing that IBM perceives that as a core competency that it needs to retain, but you're talking about working with partners who have better scale and expertise. Why not just let somebody else design or at least manufacture them?
Sanford: In that particular case, we believe we do have a core competency in the design of these microprocessors, and the learning and the improvement of that design comes from being able to manufacture it. So while we haven't given up to others the actual manufacturing of these processors, we have brought in other partners that work with us on the manufacturing. It helps to share again expertise and knowledge. We just recently announced that we re-upped with Sony and Toshiba (for the Cell processor).

You talk in the book about the connection between commodities and standards. When something standardizes, that helps commoditize the market. But it seems like a lot of the things you're talking about are very far from a standard--HR processes, for example. Once you have the standards, you can shuffle functions from in-house to out-of-house, or from one out-of-house operation to another, but it seems to be that these processes are far from standardized.
Sanford: Not everything is standardized, you're absolutely right. Some are further along in that process. In my book I talk about Magna Steyr, who is doing a lot of the design work for BMW. As they bring in more and more work from many other partners around the world, it gives them the economy of scale or the economy of expertise that then leads to standardization. If you look at a FedEx or UPS, they do logistics for people, they are defining standards as they go along in the processes that enable them to bring in someone else's logistics. And in IBM, we have a very standard way in which we run IT shops that allows us to bring it in as well as move it out when companies and customers want to take it back out again.

That's an interesting point. Recently (Sun Chief Executive) Scott McNealy has been on a tear complaining about the switching cost, the "barrier to exit," for companies trying to move off of Windows PCs and IBM mainframes. Do you think that it's actually pretty easy to switch off of an IBM mainframe?
Sanford: I do think that a lot of what we had done over the years was to allow (the mainframe) into an open server environment. For example, you can run a partition with Linux and therefore get the portability of applications from one environment to another. But to your point, I do think there needs to be a lot more standardization happening. Quite honestly, sometimes the biggest sinners are the best repenters, and I think we learned our lesson. While we enjoyed it in its heyday, at the end of the day, it probably inhibits you from the growth you're looking to achieve in a sustainable way. So we are, as strong as we were on the mainframe way back, we are equally strong now on open industry standards.

So what is the role of proprietary lock-in today? A world of standards and commodities is contrary to the idea of proprietary lock-in, but you argue companies should adding proprietary technology or products or processes on top of standard stuff. It certainly seems to me proprietary lock-in is a very important part of a lot of business models.
Sanford: This space is being rethought right now, and IBM is in forefront of this issue with this potential conundrum that you've described. The world more and more believes that innovation in the 21st century is going to come from collaboration. It won't necessarily reside all within the four walls of the single company or within the head of a single individual. How do you encourage and create an environment where both collaborative innovation and proprietary innovation can co-exist? Because while you need to connect things and collaborate in an open way, you still need to differentiate yourself. You need to find the right balance. The way we've looked at it is to look at some base capabilities being defined through the open, standards approach with collaborative innovation. Companies who then build on top of that open base bring their own value-add that will help differentiate them in the marketplace.

Do you think the world of proprietary lock-in is relevant, or is it doomed to fall by the wayside?
Sanford: No. But the reason I use the term value-add (instead of "proprietary") is because I think quite honestly if it is value-add, whether it's open or proprietary, customers are going to go for it.

Let me ask a related question then. Microsoft Windows: Is that proprietary or a standard? To a certain extent it's a standard because it's very widely used. There's a large ecosystem of other technology that goes along with it because people can safely assume that Windows is going to be there on PC. But on the other hand, it's not a standard in the sense of somebody else being able to offer Windows the way, for example, you can get x86 processors from Intel or from AMD.
Sanford: Let me go back maybe a step or two to define innovation, because I think that's important here. I believe innovation is a marriage of invention--the next new techno-whiz thing--plus insight. It's the application of technology. The theme I write about is more on how you enable innovation in a way that delivers more value to your customers. Innovation in the past had been thought of more and equated to invention--the next new iPod, the next new Mac, the next new Windows operating system. I don't think it's going to come from there. It's going to come from how we apply technology to reinvent what we do.

So what would your advice be for Microsoft in dealing with Windows? Is that something that they should let go of? A lot of what you're saying certainly sounds good in theory, but what would you advise Microsoft to do itself?
Sanford: What I'm advising our clients on is taking a look at their core business, where the value-add is, and what they need to continue to do themselves, versus what they need to partner with others on. It's not so much advising a technology company on how to necessarily focus on proprietary versus opening. Obviously IBM supports open industry standards--we are a leader in that initiative. Microsoft's got to decide what Microsoft's got to do based on what they see as a market opportunity--and not only based on history, but more on the future.