Speeches

ANN LIVERMORE
Results-driven outsourcing: balancing the risk and the rewards
Gartner Outsourcing Summit
Las Vegas, Nevada
May 18, 2004

Good morning.

Whenever I talk to customers these days, all of them are focused on getting a better return from their IT investment. It doesn't matter what size company they are, or what their industry is. Every single one of them is thinking about how to get a greater return from their IT investment.

I know that some have argued recently that IT doesn't matter, because everybody has it. But that's an argument based on quantity, not quality. Just because everyone has IT, doesn't mean that everybody's IT is the same. The businesses that are winning in the marketplace - and will win in the marketplace - are those who understand that there is no separation between information technology and business.

In fact, I would argue that technology and innovation are more, not less, important than ever. The opportunity to apply technology to fundamental business and social issues is enormous. Think about some of the issues that we have today in fundamental areas like healthcare, education, cyber-security.

To a large extent, these issues will depend on technology in order to be solved. Therefore, the opportunity to turn innovation into competitive advantage is huge.

It may be true that businesses are spending less on technology today than they did in the late 90s. But it is also true that business executives understand technology much better today than they did then.

You are no longer willing for IT to be a science experiment. You want technology to be tracked and measured, just like every other part of the business. You want IT to have a clear and measurable return that everybody can understand, just like every other business decision and business investment that gets made.

For those of us in technology, we need to adapt the value and delivery of IT to the demands of the business.

And not only are business needs changing, but technology itself is changing. I can point to three "big shifts" that are happening right now: fundamental changes that are affecting each and every one of us not only in our businesses but in our lives, as well.

The three big shifts

Everywhere we look, it's increasingly clear that we are entering an era where every process and all content is becoming digital, mobile and virtual.

For an easy example, just look at photography today. Photography used to be a physical, chemical and analog process. Today, you create digital content. A digital camera is basically a computer with a lens. And then you take that content, and you network it, you send it wirelessly, you edit it, you share it - and then, when you are ready, you print it.

The same thing is happening with music and movies. It's happening in healthcare, government and education. It's even happening in the manufacturing supply chain, where HP is one of the leading companies involved in RFID, or radio frequency identification. This technology, which allows for real-time inventory tracking, is being implemented in businesses like Wal-Mart as well as in public entities like the U.S. Department of Defense.

So every process will become digital, mobile and virtual.

The second big shift is that as technology becomes more embedded in every aspect of business and life, there is an increasing demand for simplicity, manageability and adaptability. From business leaders to consumers, people simply want things to work better together.

And the third big shift is that the world is becoming horizontal, heterogeneous and networked. Whether you're a CEO who is trying to become more efficient or a consumer who is trying to get a bunch of devices to work together, today's challenge is all about horizontal connections.

It is about how technologies work together, how networks work together, how systems work together, how applications work together and, ultimately, how all of these things enable people to work together.

This is an area where standards are important. Standards are important because they represent a common language.

Now, there are some companies who think that standards are only about lowering costs. And it is true that standards can drive down costs for clients. But the real business value is that standards can create common languages, so that different applications, different processes, different systems and different platforms can talk together.

So, processes and content are becoming more digital, mobile and virtual. Technology is becoming more simple, manageable and adaptable. And the world itself is becoming more horizontal, heterogeneous and networked.

The delivery model of the future

Information technology is changing. The way that we use information technology is changing. And within enterprise companies, the way that IT organizations deliver technology to their business partners - the "delivery model of the future" - will determine which companies get the best return on their IT investment and which companies win in the marketplace.

So, what will the delivery model of the future look like, three years from now? For the rest of my time this morning, I would like to talk about what I see coming.

From the strategic work that we're doing with our customers and partners to the investments and decisions that we're making inside of HP, I'd like to share my perspectives on what lies ahead.

The delivery model of the future will be characterized by six major attributes. And I'll go over each of them in detail.

IT and business synchronized

The first attribute of the delivery model of the future is that IT will be much more tightly aligned with the business. At the ideal end-state, business and IT will be completely synchronized, to capitalize on change.

This is clearly an area where a lot of companies have an opportunity to evolve. As part of HP's own Agility Assessment Service, which helps our enterprise customers to measure how adaptive they are, we established a benchmark of 300 large companies, across various industries.

When we asked them which areas were the least agile, one of the most frequently cited was, "adjusting IT budgets to changing business conditions." This was true across just about every industry we examined.

Bob Napier, who was HP's CIO during our huge merger with Compaq, which was, by the way, the biggest merger in IT history, was very fond of saying, "Every business decision triggers an IT action." Think about it. Every business decision triggers an IT action.

So in this future state, business and IT will be synchronized. What do I mean by that? One way to look at it is to describe it as a journey from where we are today.

At the lowest end of the spectrum, IT is very technology-centric. The strategy is all about control - getting in control, staying in control, maintaining control. And frankly, at this first stage, business pretty much views IT as a problem. One more thing to worry about.

So the journey to business-IT synchronization begins with a transition. In this second stage, IT's strategy is now to get close to the business. And business begins to view IT as a solution.

At the third stage of this transition, IT becomes business-centric. The strategy becomes all about getting IT into a service culture and delivering IT as a service. At this stage, business begins to view IT as a partner.

Finally, at the end of the journey in the fourth stage, the ultimate goal of IT is to become completely user-centric. Here, IT's strategy is to get into a leadership position in helping the business to win in the marketplace. And in this state of absolute synchronization, business views IT as an opportunity.

There are many names for this journey, and this ultimate end-state. At HP, we call it the Adaptive Enterprise: business and IT perfectly synchronized, to capitalize on change.

And as you get more sophisticated, as you align your business and your IT more tightly, one key element of a well-run enterprise is effective business-IT governance. And by the way, this is also a critical element for any effective long-term managed services partnership.

To give you an example, let me share the governance model that HP uses in our engagements. Our IT-business governance model is comprised of four main functional pillars. And these are supported, of course, by a variety of tools and enabling processes.

The first governance pillar is strong relationship management. You need to have linkages all the way up to the executive level in order to ensure that you are aligned on the big picture. And at the same time, you need to map the key contacts in business and IT all the way down to a tactical and project management level.

The second governance pillar is operational control. This includes the basics that everyone expects. Are the services being delivered on time? Are the systems up? In an outsourcing relationship, are the service level objectives being met? Basically, operational control ensures that the trains are running on time.

The third pillar of governance is demand management. This is where you must link the drivers of demand in the business organization with the cost implications in the IT organization. And honestly, this is the area where most relationships get into trouble.

Remember the quote I shared from Bob Napier: "Every business decision triggers an IT action." As business decisions are being made, which drive cost into the IT equation, are the people who are driving those costs accountable?

Let me give you a real-world example. I'm not going to mention any names, but it will be obvious to most of you who I'm talking about. There was a large service provider, who contracted with a large IT shop in the U.S. government. Part of this outsourcing agreement included configuring a user PC for every client.

The trouble was, the service provider was not paid until the client accepted delivery. So there wasn't clear accountability on the client end. For example, several users would already have a desktop PC built and configured for them but they would decide they wanted a notebook instead. So they would order a notebook PC.

This kept driving cost into the model, but there was no client accountability. The service provider had to eat it. As you can see, there was a fundamental problem with how the deal was structured.

So demand management is a critical pillar for successful business and IT governance.

The fourth and last governance pillar is strategic IT planning. There needs to be continued interaction to support future business activities. Business and IT must stay linked, in terms of how IT will enable future business strategies.

So this is one example of a governance model for business and IT. This is absolutely critical, whether you are talking about an outsourcing model or the way that IT works with its business partner within the enterprise. Synchronization between business and IT is the first attribute of the delivery model of the future.

Adaptive

The second attribute of the future IT delivery model is that it will be adaptive. IT will be built with change being the norm, and it will be built to capitalize on change.

Today, a disproportionate amount of IT is tied up in maintaining the status quo. In fact, our customers tell us that a full 42 percent of their IT budgets are devoted to maintaining the infrastructure, while only 23 percent is devoted to areas like application innovation.

Almost half of the budget is devoted to maintenance. Obviously, many companies are looking for a way to reverse these numbers: to decrease the infrastructure maintenance and increase the percentage of the IT budget that can be focused on innovation and, in turn, business agility.

How do you build a delivery environment where change is the norm? One key is to automate the dynamic link that exists between business and IT.

Again, I will use HP's approach as an example. In order to enable an Adaptive Enterprise, we use a reference architecture that we call "Darwin."

As you may recall, it was Charles Darwin who said, "It is not the strongest of the species that survives, nor the most intelligent, but the one most responsive to change." And this is true whether you're a species trying to thrive in nature or an enterprise company trying to thrive in today's competitive and changing business environment.

So we've created the Darwin Reference Architecture. And the Darwin Reference Architecture is comprised of several different layers. At the top are the business processes. Supporting the business processes underneath are the applications.

And supporting the business processes and the applications underneath is the IT infrastructure. Then across all three layers - the business processes, the applications and the IT infrastructure - you have the management piece.

So, to build an environment that is adaptive to change, you need to automate the link between business and IT. You need to optimize the business processes and applications. You need to optimize the infrastructure underneath. And you need to optimize the management that links them together.

In our Adaptive Enterprise approach, we use four overall design principles in our methodology: simplification, standardization, modularization and integration.

Simplification is all about reducing complexity and risks, by consolidating on fewer components.

Standardization is about increasing flexibility through standards-based components and processes as well as leveraging best practices.

Modularity gives you the ability to change, manage, virtualize and use components both collectively and independently. It's all about getting the maximum utility out of every single asset. For instance, many of you tell us that today, you use only 20 to 40 percent of your IT infrastructure. That utilization rate ought to be 80 percent or more.

And lastly, integration is all about putting all of these components together into a uniform system that is easy to manage, modify and change. This is true whether you're talking about data, application and process integration across the enterprise or integration across an entire ecosystem with clients and partners.

These four design principles - simplification, standardization, modularity and integration - can be applied throughout the Darwin Reference Architecture. They apply to the business process layer. They apply to the application layer. And they apply to the infrastructure layer.

So as you can see, we're slowly building the attributes of the delivery model of the future. The first attribute is that business and IT will be synchronized. The second attribute is that the model will be adaptive, to capitalize on change.

IT delivered as a service

The third attribute of this future IT delivery model is one that I alluded to earlier. IT will be delivered as a service.

In most of today's enterprises, IT is still very functionally oriented and IT charges are very much centered around assets. We see charges based on the number of servers or the number of desktops or the amount of storage space used. We see bill-backs according to data center square footage. And if you think about it, these asset-based charges are pretty meaningless to the businesses that are getting them.

IT needs to shift from an SLA, or "Service Level Agreement" approach… to an SLO, or "Service Level Objective" approach. SLAs measure basic areas, like the time to respond, the time to repair and the percentage of uptime.

SLOs, on the other hand, measure areas that are much more meaningful to the business: things like the availability of the service, deployment metrics, change metrics and overall user satisfaction.

In other words, IT needs to evolve from being concerned with technology outcomes to being concerned with business outcomes.

Instead of technology outcomes, like platform or network availability, future IT delivery will measure its success based on business outcomes, like business process availability and business function availability. Not just are the servers up, but is the application available, as well as everything that needs to support that application for the client.

This is an area where there's a lot of evolution. Today, it's still an emerging area. Because the only things that get improved are the things that can be measured. And it's sometimes difficult to determine what aspects of a particular business process can be measured.

At HP, we use methodologies that are based on ITIL standards. ITIL is a worldwide standard for the provision of IT services, and it forms the basis of our own Service Management offerings.

So the delivery of IT needs to be managed as an end-to-end process, around services that deliver what the business needs. As we take this to the next level, we are viewing IT delivery as a supply chain. A seamless, end-to-end process from the moment that a service request is made until it is fulfilled.

As the links are tightly coupled and increasingly automated the business sees the benefits that are typical of any optimized supply chain: low cost, high quality, high velocity and repeatable processes.

As in any other supply chain, there's an order management function. There's resource rationalization. There's a build and manufacture function. And there's a distribution or fulfillment function.

And underlying the overall process, you have several supporting capabilities: things like service development, supplier management, quality management, demand planning, inventory management and partner management.

So in the model of the future, IT will be delivered as a service. And this delivery will be managed and measured just like any other business process. By viewing IT as an end-to-end supply chain, you can facilitate the supply/demand balance with the needs of the business.

Multi-sourced

The fourth attribute of the future delivery model is that IT will be multi-sourced.

Increasingly, businesses are using a variety of sourcing options. I would identify three overall categories of IT sourcing.

The first category, I would characterize as "fulfillment." This is where the client says to the provider, "supply me with industry-leading products, so that I can manage my own IT services." In this situation, clients want to build their IT themselves, they want to run it themselves, they want to develop it over time themselves. In a vendor relationship, these clients are primarily looking for suppliers.

The second category is "enablement." Here, the client is saying, "Help me to design and implement an environment, so that I can mange my own IT services." In this case, the clients still want to have some control and personal involvement in IT. But they doesn't necessarily want to staff it all themselves.

They may lack, or they may not want to manage, all of the skill sets that are necessary. So the vendor relationship that these clients are looking for is more like a consultant or systems integrator.

And finally, the third category of sourcing is "management." This can range from the tactical - where the client says, "Manage these specific IT services for me" - all the way to the strategic - where the client says, "Take full ownership and responsibility for the people, the assets and the delivery of IT services for my enterprise."

In other words, "Have somebody else take on the accountability for transforming my IT and operate it for me."

Increasingly, there's going to be a mix between these three categories of fulfillment, enablement and management. At HP, many of our large enterprise clients already have a mix of these. In fact, one of the reasons that HP has been so successful in strategic outsourcing is because of our collaborative and flexible approach.

We have absolutely discovered that one size does not fit all. So we work with our customers to find the level of engagement that is appropriate for their situation.

The sourcing engagement that a particular company chooses will depend on the strategic focus and core competency of the business. It will depend on their particular access to skills and geographic reach. It will depend on the immediacy of financial impact - how quickly they need to see financial payback. And it will depend on their ability to realize economies of scale.

So as far as IT sourcing goes, one size will not fit all. And in the delivery model of the future, IT will be multi-sourced.

Globally optimized

The fifth attribute of the future delivery model is that IT will be globally optimized.

This does not mean "offshore" or "onshore." This means mapping the globally available delivery resources with the needs of the global enterprise.

Increasingly, enterprise businesses are global businesses. They expect a single view of delivery wherever they are. There needs to be a single approach across global operations. And so, IT needs to be able to deliver that single view to its business customers.

As the world economy becomes more global, there are increasing pools of very skilled and very talented people in countries around the world. These skill sets have been available in developed countries but they are also increasingly becoming available in developing nations.

The key is to have your IT personnel in the appropriate places. You need to have some people close to the business customer for those times when it's required, either for rapid hands-on response or according to customer preferences.

But for those services that do not need to be onsite or local, there's an opportunity to achieve economies of scale by centralizing common functions.

And in order to bring this network of global resources effectively to solve the client's problem, you also need to have globally consistent delivery processes supported by a high degree of automation.

At HP, we find that a global services delivery capability is absolutely critical for us to provide both superior services and a superior total customer experience. We are a global company, and our customers are global companies.

Ultimately, IT's goal, again for the benefit of the business, is to provide high-quality service with low-cost delivery. So in order to succeed, a globally optimized delivery model requires tightly linked global processes, as well as the flexibility to be able to dynamically adjust as the business strategy and marketplace evolve.

So far, I've talked about five attributes of the delivery model of the future. Business and IT synchronized. Everything built to capitalize on change. IT delivered as a service. Multi-sourced. And globally optimized.

I have also repeated several key phrases throughout my discussion. I've referred to this future delivery model as a "journey." And I've referred to it as an "evolution."

These phrases will be important to keep in mind, as I talk about the sixth - and last - attribute of the delivery model of the future. The continued evolution of service oriented IT toward a utility model.

Evolution to utility model

The utility model has captured people's imagination. It promises a huge potential for achieving economies of scale in IT operations. This would be a valuable tool for driving down the operations component of IT budgets and for enabling more resources to be available for investing in innovation.

When we look at the IT landscape today, when we talk to customers, we see three models of utility computing that are beginning to emerge.

The first is what I would call a "private utility." This is basically a "one-to-one" relationship. A utility is built out within a firm, for use by its own business units. It can be run by a central IT organization, or it can be outsourced to an external provider. But it is a dedicated environment, with flexibility in scaling and transparency in usage.

The second model is that of a "shared utility." This is basically a "one-to-few" relationship. This can be considered as an interim stage in the journey to utility computing, where computing is shared across a multiple number of clients.

This is where you see most of the leading-edge work going on today, as far as the evolution of utility computing. This is where you have large outsourcers - including HP - acting as service providers.

The issues that we are dealing with include how to leverage across different customer environments. How to standardize the IT environment - the data center, personnel and storage.

In this shared utility model, most of the pay-per use-occurs within specific bands and ranges. For example, X dollars per user per month, within a range of Y to Z.

Then finally, the third model is a full-blown "public utility." This is basically a "one-to-many" relationship, much like a traditional utility or the original ambitions of the ASP model. Just as with electricity or mobile phones, there is a shared environment on a large scale. Providers offer a catalog of services, using a "pay-as-you-go" pricing scheme. Customers can buy services, quickly connect and switch suppliers.

Obviously, this does not exist yet. And the road to a public utility model will very much be an evolution, not a revolution. Anyone who says otherwise is either making an unrealistic promise or they're just not looking at the marketplace.

Quite simply, we have not yet achieved the correct balance between the supply side and the demand side to create an efficient market.

A good analogy is the situation with electric cars over the last several years. Electric cars could only travel about 50 miles or so before they needed be charged. The providers didn't want to create a massive infrastructure of charging stations until there were enough people using electric cars. And customers didn't want to use electric cars until there was a massive infrastructure to support them.

There is a similar struggle between supply and demand, when you look at computing as a public utility.

On the demand side, it's all about achieving risk-adjusted business value. This is a huge issue. It's a new model. And there would need to be a compelling advantage for a critical mass of firms to sign on board.

We don't need to look any further than the people in this room. What would it take for you to adopt this in a big way?

So from the demand side, it's a question of risk vs. business value. There are questions of security and privacy, questions of control, questions of confidence, credibility and overall comfort.

On the supply side, it's all about achieving an economically attractive business model. It's like the electric car example. Providers need to balance the cost of client connections with the flexibility of client switching. Providers need to deliver economies of scale while avoiding the trap of becoming commoditized. That's what happened in the storage market.

There are other considerations as well. For instance, computing as a utility is not much use unless software also evolves to a pay-per-use model as well. So there would need to be an evolution in software licensing models.

Overall, the industry is very anxious to avoid a repeat of the ASP market several years ago. As many of you recall, there were several "pie in the sky" promises made, that turned out to be overly optimistic. There was a general feeling that "if you build it, they will come."

So the journey to utility computing will very much be an evolution, not a revolution. And in the meantime, as we continue to work through the issues, you see a lot of work going on in interim steps like the shared utility model.

So, let's review. I've talked about how today's new era of technology is being driven by three "big shifts." First, everything will become digital, mobile and virtual. Second, customers are demanding that their technology be simpler, more manageable and more adaptive. And third, the world is becoming horizontal, heterogeneous and networked.

In this new era, technology is the enabler, rather than the driver, of change. And so, the delivery model of the future will have six fundamental attributes.

Business and IT will be synchronized, as well as more adaptive, to capitalize on change. IT will be delivered as a service. It will be multi-sourced and globally optimized. And while IT will move toward a utility computing model, that journey will be an evolutionary one, not a revolutionary one.

As I said at the beginning of my presentation, this delivery model of the future is driving many of the investments and decisions that we're making inside of HP.

We have organized ourselves and based our strategy around the vision of the Adaptive Enterprise. We have invested in research and development around the Darwin Reference Architecture.

We believe in this model, because we have lived through its journey and its evolution ourselves.

As I mentioned, two years ago we undertook the biggest merger in IT history, with HP and Compaq. When we brought these two companies together, we had more than 7,000 applications and 160,00 employees in 167 countries. We became the largest transactor of e-commerce in the world. We had 260,000 total e-mail addresses.

And using our own Darwin Reference Architecture, these tools, these disciplines, our products, our people, our services and our partners, we were able to pull all of that together and do business as one company on Day One.

That included an integrated Web site to our customers, integrated supply chains and integrated e-mail, so that all of our employees could collaborate, find one another and talk with one another on Day One.

And in the process of doing that, we reduced our IT costs by 25 percent. We had aimed for $2.5 billion in savings within two years. We achieved $3.5 billion in savings within one year.

And we're not done yet. We know that there's still even more that we can do. As I've been saying, it's a journey and it's an evolutionary process.

But at HP, we are successfully building our own Adaptive Enterprise. So when we talk about the delivery model of the future, we are able to speak with some experience.

In the end, the successful delivery of IT is all about creating sustainable and measurable business value. It's all about creating a model where change becomes an enabler and a competitive advantage. As Charles Darwin said, more than a hundred years ago, "It is not the strongest of the species that survives, nor the most intelligent, but the one most responsive to change."

Thank you.