Speeches

A CONVERSATION WITH CARLY FIORINA - FORBES 5TH ANNUAL CIO FORUM
DECEMBER 2, 2003
DALLAS, TEXAS

© Copyright 2003 Hewlett-Packard Development Company, L.P
All rights reserved. Do not use without written permission from HP.

Quentin Hardy:
Is what Jeff (Immelt) said seem alright to you? Is the productivity gain all about IT?

Carly Fiorina:
Yes. I have met quite extensively with Jeff and his senior management team, and like so many companies, GE has invested in IT over the last set of decades around vertical applications, separate divisions, separate departments. And now, to unlock the next wave of productivity that Jeff and many other CEOs are talking about, you have to start thinking about processes as horizontal across the company - a requirement for processes to work together better, for applications to work together better, for systems to work together better, and so it is now all about not just simply automating processes, it's about digitizing processes, making them simpler and easier to manage, and that's how productivity will be driven. IT becomes more mission critical, more complex, and the CIO therefore becomes also more mission critical, and a more, frankly in my judgment, strategic role.

Quentin Hardy:
So are you selling IT that is a box, or are you selling consulting, or are you selling services? How do you begin to approach the problem? What's the best way to think about this?

Carly Fiorina:
First, I have to correct one of the things you said in your intro. We are building a very successful professional services business, but consulting is the smallest piece of it. We are not building a big consulting business. We passed buying PWC many times, including two weeks before IBM bought them, and I could have bought them for $500 million less. The reason I didn't, and it's relevant to this discussion, is because one of the things that I've definitely seen happen with CEOs and CIOs is that they are a lot less inclined to spend money on long consulting engagements that are open-ended, in many cases, vertically oriented around specialized applications and processes. And what they're much more likely to focus on is, 'I want the player who sells me IT to help me understand the linkage between my business objective and how IT can enable that business objective; or, frankly, how my current IT systems are getting in the way of my business objectives.' In other words, that's not a long, expensive consulting engagement. That's fundamental to what we do and our customers do.

Quentin Hardy:
What I meant by the consulting wasn't so much around domain expertise as about process, and identifying how the network is going to operate and where the intelligence should reside - or is that just a subset of what happens when you sell a server now?

Carly Fiorina:
Well first, I do think - and I've said this publicly many times - I think the hot box era is over and has been over for quite a while, because all of you who know technology well and spend a lot of money on it and time worrying about it, know that a hot box alone is not enough.

Fundamentally, now it is about managing the total cost of ownership of IT - which certainly includes hardware - but also includes the management of that hardware every day, the management of the inventory, the management of the applications. You require reliable innovation at a price you can afford, because you can't have your business falling behind. You require increased manageability, increased simplicity. You require your IT systems to provide not just productivity, but adaptability. You don't want your IT to prevent you from moving and changing your business when you need to.

All those things are required. That's a lot more than a hot box, and it's a lot more than simply who's got the cheaper hardware.

Quentin Hardy:
Well, what do the CIOs worry about the most where this is concerned? How to begin? How to even think about the problem?

Carly Fiorina:
I think most CIOs that I talk to, and I've talked with a lot of you actually - I see some friendly faces out there - I think most CIOs start with the recognition that their list of demands today is as long as it's ever been, and they cannot compromise on that list. What I mean by that is, it is no longer possible to compromise between innovation or price. You need reliable innovation and you need a price that's right, and frankly, what that means is your total cost of ownership continues to decline over time. CIOs can no longer trade off between quality, reliability, and speed and flexibility - that's a tradeoff CIOs used to be willing to make. You know, 'I'll trade off quality and reliability so I can move faster and be more flexible.' CIOs now say, 'You know what? I can't trade those off any more. I need quality, I need reliability, and I need speed and flexibility.' I think CIOs also know, for example, they can't trade off any more between functionality and manageability. 'I need the functionality, and I need to be able to manage it.' So that's a big deal first, that the list of requirements is long, it is additive, and it is not subject to compromise.

I think the second thing that CIOs then worry about is, let's face it, one of the reasons I think the survey says the budgets are only going up 1% in 2004 is because most CIOs know there is still too much technology in some companies in the wrong places, and not enough technology in other places. In other words, while the total spend needs to change, the total spend doesn't necessarily need to go up. Most CIOs and CEOs that I talk to are probably utilizing, really utilizing 40% of their IT spend.

Quentin Hardy:
So, I see flat budget. I don't have to think being a CIO is strictly a defensive game?

Carly Fiorina:
No. You know, I noticed you have a panel later about eating your own dog food - Cisco and Intel eating their own dog food - and we really believe that too. One of the things that we did throughout the merger integration process was eat our own dog food in terms of our systems, and our people, and our software, and our capabilities, and we ended up driving huge IT productivity improvements - I'll give you an example in just a moment - and we reduced our IT budget by 25%.

Now, part of that was because we were collapsing redundant systems, but I guess my point is that there is a lot of overall spend that can be mined in an IT budget by spending the money differently. In our case, for example, we reduced the time it takes to add a new supply chain partner to our global supply chain from five weeks to two hours. At the same time, we reduced the number of total supply chains we use in the business from 20-plus to five. So, it's an example of thinking about processes more holistically across the business - so, we didn't need 23 supply chains, we only needed five - recognizing that in a supply chain like ours, that's worth $50 billion a year; it's the largest supply chain in the IT industry - that when you can reduce the time it takes to add a supply chain partner from five weeks to two hours, that's a huge deal.

Quentin Hardy:
That's an enormous leap going from 23 to five. How do you know you got it right? And you went to five - you didn't sort of go to 15 and see how that was, and then 10 to see how that was - you had to make the move whole. How do you do that and know you've got it right?

Carly Fiorina:
Well, I think it first starts with process architecture, and this is how we operate in our own business; it's how we work with CIOs and CEOs to operate in their business. Process architecture is all about looking deeply at what processes do you really need to accomplish what objectives. It turns out that many companies - I mean, most companies in the 90s were very vertically oriented, and in service of speed and flexibility, a lot of companies built a process for every division, an application for every division, and an IT system for every application and every process.

Quentin Hardy:
So specifically, marketing would have all its tools, and Finance would have all its tools?

Carly Fiorina:
Absolutely, so you'd have a lot of risk going on. And now, when you start to step back and say what's common about the 23 supply chain processes we have in our business - after that type of process architecture and business objective setting, we concluded in fact there are five supply chains we need, not 23. And by the way - you all are CIOs, you know this - while I just said the easy part is the analysis, the hard part is the change management, because everybody agrees with it on the piece of paper until it's time to say, 'You mean I have to change my supply chain? I have to give up control over my IT system?' - that's when it becomes very difficult, and so one of the other comments I would make - and all these guys know it - a CIO is not only, in my judgment, a business person who has to have a seat at the business table and understand what are the business's objectives? What gets in the way of the business achieving those objectives? How can IT enable - or how does IT encumber - the achievement of those objectives? But a CIO has to really understand the complexity and sophistication of change management, because change management, as well adaptability, can be enabled by technology, or technology can really get in the way.

Quentin Hardy:
Let's spend a minute more on your own case study. Your own relations with your CIO - what did you find worked really well, what were the sort of sticking points that were slowing down productivity?

Carly Fiorina:
Well I think first, the CIO of HP has a seat at my table. So, when I and the senior team of people who report directly to me meet for our, just as an example, two day-a-month Executive Council meeting, the CIO sits at that table and engages in that conversation to the same degree that any other member of that table does.

Quentin Hardy:
Had it always been that way with you in HP?

Carly Fiorina:
Well, when I first came to HP in 1999, we had 87 different product divisions and every one had their own CIO, and every one had their own system. I mean, that was a case where we were spending lots of money on technology, and none of the systems worked particularly well together. It was aggressive decentralization, which served a purpose at a point in time, but it was an organizational logic that didn't work any more. So we have come through massive change.

And one of the first things we did prior to the merger was say we are going to drive rationalization of our business from 87 different product divisions - we got down to 12 - before we undertook the merger. We're going to drive that rationalization of the business by picking two processes first, and using IT to enable those processes. And the two processes we picked initially, not surprisingly, were finance and HR. And we literally used those processes to bring the company together, and then we built technology to enable those processes to work horizontally across the company. All that was pre-merger.

Quentin Hardy:
You chose those two because they had the most readily common language, or there was the best existing base of software? What drove that?

Carly Fiorina:
No, we chose those two because they clearly impacted every part of the business, and frankly, it was the easiest place to start from a change management point of view, because virtually everyone could embrace and accept that there was a common requirement for finance and HR. So, when I would say to executives, "You know, we have over a thousand internal Web sites for employee training. How does that help our employees?" People kind of got that. And we then moved - again pre-merger - to a single @HP portal, a single technology platform to enable a single HR process across our entire company - 176 countries. And that technology platform, just as an example, allowed us on day one to bring all these new employees into the company, and instantly, everybody could communicate, everybody could collaborate, everybody could e-mail, everybody could find each other. That was an example of a technology platform enabling rapid change and accelerating integration. It's also an example of using technology to drive a horizontal process across the business that changed how we operated.

Quentin Hardy:
By the way, I'm kind of curious. Is it easy or hard to impose technology on a company full of engineers? Because I get the PC - ok, it's the PC; I get the network - I agree with the network. But over at HP, it probably looks like math camp - they all know the answer and they all want to say it, right?

Carly Fiorina:
I don't know. I've been in technology-based companies full of engineers for the last 20 years, so it's hard for me to step outside that, but I would say that it's hard in a company full of engineers, because change is always difficult. Change management is always a challenge, but for engineers, technology really is incredibly emotional…

Quentin Hardy:
But how do you impose a change to make it work? Is it about setting perceptions? Is it about having a sort of clear hard and fast rules?

Carly Fiorina:
Well first of all, I don't think you impose change. That doesn't work. I think-

Quentin Hardy:
Effect change.

Carly Fiorina:
Yes, I think effect change is a better term. I think change requires critical mass of support - which is different than unanimity of support. A required change never has unanimity of support, and by the time it does, it's usually too late. I would argue Kodak is a great example of a company choosing to do something when it was clear to absolutely everybody that something had to be done. And generally speaking, when it's absolutely clear to everyone that something dramatic has to be done, you've lost a ton of time. So, that's why I think change requires critical mass and not unanimity.

I think change as well requires alignment around objective, and it's interesting to me how many people sometimes miss that first step. If you don't have alignment on objective, purpose, goal, strategy, what are we trying to accomplish with this change, you'll never make it. Because as the change gets tough, people want to stop, and the only way to keep people going is to go back to the fundamental purpose and objective around which people aligned, and, 'Oh yeah, that's why we're doing this.' So I do think effecting change requires a clear, quantifiable set of objectives that people can align behind.

Quentin Hardy:
If you can't measure it, you've got to wonder why you're doing this as a business anyway.

Carly Fiorina:
I think that's right, and by the way, I think that's what a lot of CEOs have learned about technology in the last couple years. It shouldn't be a science experiment; it's not a mystery. It ought to be measured the same way any other business investment is measured, and that is, you've got to have numbers, you've got to have payoff, you've got to have results, and a CEO needs to be able to understand that. And I think there was a period of time when people who understood technology really well could come in and say, 'Wow, you don't get it, but trust me it's going to work.' And I think those days are over.

Quentin Hardy:
Let me ask you about change of another sort. How do you know when to outsource?

Carly Fiorina:
Frankly, I don't think outsourcing is the objective in itself; I think outsourcing is a means to an end. Our outsourcing business is growing at four times the market rate today. A year ago, our outsourced business was growing about 15%. In the quarter we just announced, it's growing at 36%, so we are seeing accelerating momentum in our outsourcing.

Quentin Hardy:
In what areas are people are outsourcing in particular?

Carly Fiorina:
People are outsourcing datacenters, they're outsourcing publications, they're now in the managed print services business where we already have 650 customers, they're outsourcing the management of their document production. But we don't advise customers, 'Go do outsourcing.' The reason I say outsourcing is a means to an end, I think customers ought to have clear objectives about what is it they're trying to accomplish in terms of total cost of ownership reduction. What do they want in terms of reliable innovation? If outsourcing can absolutely help customers achieve objectives of improved total cost of ownership, greater manageability, greater agility and flexibility, but it is a means to an end.

Some customers approach outsourcing initially by starting with something they can get their hands around like the desktop. 'Let me outsource the management of the desktop, because even though I've driven down the price of the hardware a lot, in the last year or two, I now realize that only 30% of the total cost of the desktop is in the hardware; the rest of it is in the management.' So, some customers start small. Some customers want to go through a consolidation on their own of the datacenter investments they've already made before they outsource. And some customers say, 'I have a very clear set of objectives here, and I want a partner who is going to stand with me throughout the life of this agreement and deliver on those objectives.'

By the way, just one other comment on this: I also tell CIOs and CEOs all the time, never, never, never outsource your CIO - never outsource the strategic connection between the business and IT. Because when we have an outsourcing engagement, we want a CIO as a partner who sits at the executive table. That's important. You can't outsource your business advantage. You can rely on a partner, but that partner has to have someone to work with whose job it is to make sure the technology and the people who provide technology deliver.

Quentin Hardy:
Let's go back to your own case study. What did you outsource in HP, and beyond that, what were the areas of surprising cost savings as you sort of revamped the corporation?

Carly Fiorina:
When we talk about outsourcing inside HP, we are talking about using our own managed services business, our own HP Managed Services, to manage our environment. So, we're not outsourcing to a third party, we're outsourcing to HP Services. That's part of what I meant when I said we eat our dog food - we use our own people, our own processes, our systems. So, HP Services manages all of our IT systems. We've outsourced our financial processes, we've outsourced a lot of our HR processes to our own folks, and yet, even though HP Services manages that on behalf of HP, my CIO still sits at the table and reports out on whether we are achieving our business objectives through the application of IT.

Quentin Hardy:
And what were some of the surprising areas of cost savings as you put it through there? I mean, you probably knew you could collapse a few supply chains into each other. What else did you learn?

Carly Fiorina:
I think to me one of the interesting and pleasant surprises in all of this, not just since the merger, but in the four and a-half years that we've been doing this, is I said at the outset that change is hard, and yes, change is hard. But it is also true that change does gather its own momentum. And what I have observed in the last four and a-half years is people start to get more and more appetite for change. It's sort of like working out at the gym - the more you do it, the more you can do it. And they also get excited about 'Gee, we've done this in HR, we've done this in finance. We can do this in supply chain.' People tend to really understand the power in starting with process architecture, starting with that long list of requirements that I went through and not being willing to compromise on those requirements, and understanding how much technology can really enable.

Quentin Hardy:
Let's take it to the broader industry level now. I want to go back to what I was talking about in the intro about this highly networked role in that a lot of intelligence throughout the system, from the edge right to the center - in general terms, this is what you're talking about with adaptive systems, right?

Carly Fiorina:
Well, our adaptive enterprise strategy and our Darwin Reference Architecture - which is the underpinnings of that adaptive enterprise strategy - fundamentally, what that is about is first, thinking about IT from the process level down, not the IT systems up. Secondly, it is about understanding that there is intelligence in all aspects of an organization, whether it's infrastructure, applications or processes, and how is all that intelligence managed and connected in a simple, reliable way that delivers lower total cost of ownership, improved risk mitigation, and at the same time, allows a business to adapt at the pace it needs to adapt.

Quentin Hardy:
And similarly, this is akin to what IBM is talking about with on demand, and Sun with N1 - why is this a key theme now?

Carly Fiorina:
I actually think they're different in the following sense: on demand is a particular technology deployment that is a subset in my view, of what an adaptive enterprise is all about. We also have - and by the way, I would say the same thing about our on demand offering - we have something we call a Utility Data Center. I think Utility Data Centers are akin to on demand in the sense that they are IT as a service. But whether or not you move to IT as a service, it is still the case that technology can enable an adaptive enterprise.

I think the reason it's a key theme is because when you asked a while ago how customers get started, fundamentally, I think there are four key things that customers need to do: they need to simplify their process architecture, their application architecture, and their IT infrastructure. They need to consolidate those same things, and you can't consolidate until you simplify. And then to the point of on demand, they need to modularize, and what modularity is all about is allowing IT resources to be flexible and to move with the requirements of the business, and that, in many cases, can be accomplished through a service-based approach as opposed to a technology-based or hardware-based approach.

So, I think this drive towards modularity, resource flexibility, is what makes Utility Data Centers or on demand a key concept. And then finally after modularize, the last step is to integrate what had been standalone processes, applications, islands of technology.

Quentin Hardy:
We were talking a little bit before the session began, and one thing that I'm kind of curious about is, in a highly networked world where some of those silos have started to join together, does risk mitigation risk management change? Don't things start to have knock-on effects outside their own system?

Carly Fiorina:
Yes, I think connected systems are an example of classic risk reward in some cases. The more connected a system is, the more valuable it is. The more supply chain partners you have in your supply chain, the more valuable that chain is. It's also true that the more connected a system, the more places there are where risk has to be managed. So, the reward is higher, and the risks that have to be thought through and managed are higher, as well. And I think sometimes we're swinging the pendulum in terms of media coverage once again to 'Oh my God, security is a major issue absolutely everywhere.' There's no question that risk mitigation is higher on people's radar screens today than it has been, and that's for many reasons. It's for September 11th, but it also is because of the increasing connectiveness of partners and systems, and frankly, the more connected things are, the more valuable they are, and also the more vulnerable they are.

Quentin Hardy:
Yes, but if Web services really gets going, it will be hard to tell where one company begins and another one ends.

Carly Fiorina:
Absolutely right, because the power now is in the connection between things - companies, applications, processes, technologies.

Quentin Hardy:
So on the one hand, you've got things better wired on a kind of global basis than ever before, but you've also got a network where there's more intelligence at the edge with more autonomy than ever before. Is it easier, or is it harder to forecast in that kind of environment?

Carly Fiorina:
See, I don't think it's true that there's more intelligence on the edge only. I think it is true that there is more intelligence everywhere, so I don't think that this is the old argument of 'is the intelligence in the center or is the intelligence at the edge?' Intelligence is everywhere, and it's about managing that intelligence.

Quentin Hardy:
Where the rules resides, do you think that the rules should reside at the center?

Carly Fiorina:
No, not necessarily. I think rules should reside where they make the most sense to reside. And I think frequently people think about process architecture and horizontal process decision-making as centralized decision-making. It's not. It's horizontal, not vertical. I mean, it is a different orientation for people. A supply chain - a horizontal supply chain across a company - is not a decentralized thing. It is a horizontal decision process, not a vertical decision process.

Quentin Hardy:
So does forecasting get easier or does forecasting get harder?

Carly Fiorina:
Depends on how you define forecasting. If forecasting is predicting trends in the future based on current information, you have a lot more current information. On the other hand, stuff happens that you're not able to see even with a ton of information, and the downturn in the economy was a clear case of that. Everybody had loads of information, and everybody missed it.

Quentin Hardy:
So the challenge is to turn the data into some kind of real insight?

Carly Fiorina:
That's right. Information and insight are not the same thing. So, more intelligence doesn't necessarily mean more insight.

Quentin Hardy:
You've got that one figured out? How do you turn it to insight?

Carly Fiorina:
Well I think there is no silver bullet, so that is the job of management in many cases. And an IT system with all the intelligence in the world can't deliver management judgment. Managers have to deliver judgment. And sometimes, that management judgment is going to be in direct opposition to what conventional wisdom says is going to happen.

Quentin Hardy:
We've got about 15 minutes left. I'd like to take questions from the audience, if anyone has questions.

Audience question:
Yes, my thoughts have revolved around government requirements, and as government requirements become more imposing in the IT environment, for example, today ranging from the White House Cyber Security Plan, which is largely a suggestive document, and then going all the way through the US Patriot Act, HIPAA and Sarbanes-Oxley, which now has an enforcement arm in Washington D.C., PCAOB. What advice can you give the CIOs that are trying to rationalize these requirements in their environment, and what is HP doing to help with that?

Carly Fiorina:
Well there's a great question. First, in some ways, I think Sarbanes-Oxley 404, which is the part of Sarbanes-Oxley, as you know, that requires companies to document their processes for the purposes of assuring process control. By the way, it is an onerous set of requirements. But in many ways, 404 is an opportunity that can be used by a company, and particularly by a CIO, to drive process orientation in the business. We have been talking about how difficult it is to get organizations to start to think - instead of vertically around separate divisions and separate products and separate applications - horizontally around processes that cut across a company or several divisions.

404 can help a company drive to that, because there is no question that controlling risk and mitigating risk requires greater simplicity and greater consolidation of processes and applications, and the underlying IT infrastructure. That's just a fact. The more different systems and processes and applications you have going on, the more difficult it is to control them - not only the costs, but the output. The more difficult it is to control them, and therefore, the more difficult it is to manage the risk associated with them.

So I do think some of these regulations, some of these government requirements, are an opportunity. In terms of 404, just as an example, one of the things we do inside our business and one of the things we do with some of our customers is, -let me just take you inside our own business-our CFO, our General Counsel, our CIO, and our process owners are absolutely joined at the hip around how will we meet these requirements. So, we have an aligned view of those sets of players. CFO, CIO, General Counsel, process owners. We have an aligned view among that community of what the process requirements are, and therefore, how are we going to tackle them. And I think that kind of alignment is critical.

I had a little bit of difficulty understanding the first part of your question - the mike wasn't that great. So you had mentioned another, I thought, government agency, but I couldn't quite hear it.

Audience question:
No, it was more along the lines of with all the issues around HIPAA compliance now. Sarbanes-Oxley combined all that with business continuity, for example, how can the CIO possibly take all these cost requirements and create a budget that's reasonable for the company, and what advice do you have for CIOs with that challenge?

Carly Fiorina:
So first, I do think that when I walk through the four steps how do you get started - simplify, consolidate, modularize, integrate - I do think that the place everyone has to start is to look at their current IT environment, and maximize every opportunity for simplification and consolidation. Because without simplification and consolidation, you can't deliver a reasonable budget, but you can't manage risk either. And, I do think that most CEOs really do understand that process complexity, application complexity, IT complexity drives cost and risk, and so I would start with: we've got to simplify and we've got to consolidate. I guess the second thing that I would say - and by the way, we have processes, systems, services that we offer customers to help them take that first step. Where are your opportunities to simplify? Where are your opportunities to consolidate?

I do think that one of the realities of business, which sometimes people who talk about Sarbanes Oxley may forget, is business still is about taking risk and managing risk. Business is not about eliminating risk. Without risk, there is no progress; without progress there is no business. So, there is no way to get away from the reality that it is about taking risks and managing risks - not eliminating risks. And that requires judgment. And in part, the judgments that have to be made are, you know what? I can't - we can't as a company - comply to the maximum ceiling of absolutely every recommendation out there, so we have to pick and choose. That's what managers get paid to do - make judgments. And I think it's increasingly what CIOs also are paid to do. Make judgments about what risks are worth taking and what risks are not worth taking.

Audience question:
Last year at this conference, your then CIO, Bob Napier, presented and he passed away in October. Many people in this audience were friends of Bob, how would you assess his performance particularly in the merger, and what were you looking for in his replacement? What was your decision process there?

Carly Fiorina:
Well, Bob was a huge loss - as a CIO, as a colleague, as a friend, as a counselor. And it is a loss that I feel deeply and his colleagues feel deeply. Bob was a great CIO, and he was a great CIO because he understood the business. He understood what the business was trying to accomplish. He viewed technology as a means to an end, not an end in itself, meaning he wasn't religious about technology. He was religious about what technology had to achieve. And technologists sometimes do get religious about technology, and it's the wrong religion.

He was tough - CIOs have to be tough. He was demanding. He was demanding of himself and his team. He didn't walk in and demand a budget he thought we couldn't afford. He would walk in and say, "Here's what else I can do to drive our costs down, here are the consequences, here are the tradeoffs, and here is where I need to make an investment." He would never come in with an investment request without also having an opportunity to save money.

He would walk away from IT projects where he felt we did not have clear business objectives associated with them, and a business owner who would stand up and say, 'I support this change, and I will support the change management required to make it work.' If he couldn't point to somebody around the table who felt like they owned it as much as he did, he said: "I don't want the money, I won't spend it, and we wont' implement it." And that is the right discipline.

And so, I looked for all of those things. And we are very fortunate in having a very deep bench. That's one of the benefits of having an IT department in an IT company; you have a lot of people who really understand it well. And we will be announcing his permanent replacement shortly, probably by the end of this week, but fortunately, we had lots of very highly qualified people to choose from.

Quentin Hardy:
Has the job changed inside HP? Does it have different requirements?

Carly Fiorina:
Oh absolutely. If you think back to what I said we started with in 1999, I mean, we didn't even have a company CIO in 1999 - we had 87 of them. Yes, it has changed. It has become a more important job. It has become a direct reporting job. Three years ago, it wasn't a direct reporting job, even though we had a single CIO for the company. It's as critical a part of our business success and how we operate the business as a CFO, a business head, or anyone else. And that absolutely is different over the last several years.

Quentin Hardy:
How do you train your young generation of managers - IT, but also generally in the business, given how highly networked the business is at this point? Do you have to train differently?

Carly Fiorina:
I do think you have to train differently. One of the things that HP traditionally did, and I think a lot of companies traditionally did, is it groomed people and promoted people on the basis primarily of subject matter expertise. So, in other words, people rose through a particular part of the business. You know, if you started in Imaging and Printing, you ended in Imaging and Printing. If you started in the PC business, you ended in the PC business. And over time, what we've learned is that we have to train and develop process orientation, broad business understanding and judgment, and so those are things that require more horizontal expertise and process orientation, not simply vertical subject matter expertise - that's one very large difference.

I think another very large difference in a company as large and complex and global as ours - you know, we're not quite $80 billion yet, but we will be soon, but let's call it $75 billion, in 176 countries, and 140,000 employees - is virtually every manager, when they reach a certain level in the business, has to have a basic understanding of the breadth and depth of the business. It's not okay to be a middle manager and not understand what the fundamental strategic purpose and operating logic of the company is.

Quentin Hardy:
Yeah, but given seven hours of production between you and Dell, one point of market share between you and IBM, you can't afford dilettantes either. Isn't that kind of like asking for everything?

Carly Fiorina:
No, but it is about being very judicious in what kinds of developmental experiences you give people. Taking people for example, who have developed rich vertical subject matter expertise, and moving them into jobs that stretch them horizontally; so, for example, we have taken people who grew up in Imaging and Printing, and made them accountable for supply chains across the business that serve more than Imaging and Printing, just as an example.

And I guess the final thing that I would say around management training, and it's something that business people don't talk about that much, and maybe CIOs, or we don't talk publicly about that much, but I think it's critically important particularly in this day and age. It's back to this issue of judgment. Judgment, what risks are worth taking, what risks aren't worth taking, ethics - you know, I call it an internal compass, but what keeps people grounded, what keeps people's rudder heading in the right direction - those are really important things.

Quentin Hardy:
You can't teach that though, can you?

Carly Fiorina:
Well, first you can emphasize it; you can make it a requirement. There are lots of smart people who don't have any judgment.

Quentin Hardy:
But that was their mom's job, you know?

Carly Fiorina:
Well, I don't agree with that. I don't mean to get to philosophic or esoteric here, but all of you know you have people in your organizations who are really smart, but whose judgment you question. And what you do with those people is either contain them, so that their judgment cannot harm the organization-

Quentin Hardy:
Look at those knowing smiles, my goodness.

Carly Fiorina:
-Or, you have to place a requirement on them and experiences to help them achieve judgment over time. And if you can't, those people become a liability, not an asset. So, I think business judgment is something managers have to worry about, they have to develop, they have to train, and they have to require.

Audience question:
My company is probably much smaller than most people in here, which may drive my question, but I do face some naysayers in the Microsoft-versus-anything else, and I have an Oracle data warehouse, and some HP that I really like, as well as, of course, some SQL servers and many things there and I have to combat that daily, and I wonder what you would suggest there?

Carly Fiorina:
And their point of view is?

Audience member:
If you look at the word 'consolidate' for example. Some people may say, 'Well great, then I want 100% Microsoft shop.'

Carly Fiorina:
So, it's interesting because while we very much preach consolidation and practice consolidation, we also think heterogeneity is here to stay. And the reason we think heterogeneity is here to stay is because of what I said earlier, it is all about the power of connections. So, you can make your own shop all Microsoft, but you can't dictate to your customers, your partners, your suppliers. It doesn't solve the simplification, consolidation, integration - you can't solve that problem by going to a homogeneous environment. You can only solve that problem by managing effectively a heterogeneous environment. By the way, that is not everyone's philosophy; it is not what everyone preaches. We think it's a big deal, for example, our OpenView management software is based upon a belief that its job is to manage heterogeneous environments.

So, the OS wars are over. Everybody won. You've got to have Linux, you're going to have probably some version of UNIX, you're going to have some Windows, and you have to be able to manage all that. It's important, I think, that as people move to Web-based services, to acknowledge that it's not going to be .Net or J2EE - it is going to be both, even if you can contain your own little environment, you cannot disconnect your company from the rest of the world, so you're going to have to have the ability to interoperate, and integrate, and connect both J2EE-based services and .Net-based services.

So, I guess what I'm saying is, I think when people equate homogeneity - a single technology, a single application - homogeneity, that's the way to achieve simplicity, or that's the way to achieve consolidation, or that's the way to achieve connection and integration. It's not the case. You may achieve it in the datacenter, but you won't achieve it with the rest of the process. The way to achieve those things is to manage effectively heterogeneity, and it can be managed. It has to be managed by technologies that know how to manage it, and companies that are equipped and prepared and committed to manage it, but it can be managed.

The other thing is, heterogeneity over time gives you greater flexibility, because there are some things that are tuned better to different applications, different environments, different opportunities than other things, which is why the OS wars are over and everybody won. Linux works better for some things, but it's not perfect for everything, as an example.

Quentin Hardy:
Okay, quick forecast. Do you think these kinds of productivity numbers are sustainable? Maybe not 8% but-

Carly Fiorina:
I'm not going to give an exact number, but what I do know is this: technology can enable productivity. Such an obvious statement, but there is technology at every process inside a business; every process in life is going to be digitized and we're seeing it happen; every physical process will become a digital process. And by the way, every physical process will not only become a digital process, it will become a mobile process and a virtual process.

And that drives huge productivity, because physical processes by their nature - I don't care if it's photography, which has been transformed from a physical process to a digital process, or whether it's supply chains. When physical processes become digital, and then mobile, and then virtual, the amount of power that is unlocked, the amount of cost that is reduced, the amount of friction and time that is taken out of the system is dramatic. And if you doubt it, all you have to do is think about your own little example as a consumer, about the difference between taking a traditional, physical print. Now, I talked about Kodak, but to me it's the simplest example to help people get their hands around it. Think about the traditional photography process: you take a picture, something physical chemical happens in the camera, you walk or drive the film someplace. You put it in an envelope. You write a bunch of things. It goes away. Something physical chemical happens. You go back. You pick it up. You bring it home. You open your prints. You throw half of them away. You put the rest of them in your bottom desk drawer, you get to them eventually. That's a physical process, and that physical process is labor intensive, time intensive, inflexible, and not very satisfying.

What is digital photography? It's a digital process now that's also gone mobile. You can take - what is it? You're creating digital bits, and you're networking them, sending them, manipulating them, editing them, storing them, and ultimately when you're ready, you transform them into a physical print. That's a totally different process that's faster, more satisfying, more ubiquitous, and it has changed the face of how consumers think about photography, and it's changed whole industries. That is happening to every process. Not all at once, not at the same pace, but it's happening to every process - healthcare, education, supply chain, photography.

Quentin Hardy:
The photography example is a great one for productivity gain. If we can do this well by cutting the budgets for a couple of years, when are we ever going to loosen up?

Carly Fiorina:
Well, there is an old saying in business that recessions sometimes are cleansing processes. They kind of focus the mind. They're tough as hell, but most of the time, the companies that exit out of them come out stronger than when they went in.

Quentin Hardy:
All right and on that, thank you very much.