A Steady Hand in Turbulent Times

by Jeroen van Tyn

Given the stunning economic events of the past few months, it comes as no surprise that the data for this year's CBR survey on IT trends reveals some significant activity across industries to cut costs and spend money wisely. Respondents across the board report taking such obvious measures as sharply cutting back hiring, reducing headcount, and taking on fewer projects.

However, as we look at trends that have developed over the past four years of the CBR trends surveys, we see that companies are also continuing to invest in strategic initiatives, such as service-oriented architecture (SOA) and enterprise architecture (EA), that can help create sustainable efficiencies and support value creation over the long term. Indeed, such strategic initiatives are tending to focus more on cost savings rather than on value creation during the economic downturn. It appears companies have gained enough pragmatic experience to understand that the commoditization of IT is a metaphor that can only be taken so far. By their nature, the operations of any robust enterprise are dependent on IT for execution. It stands to reason that if businesses are to survive, they will need to have the stomach for the long haul; strategic IT investments are an integral part of that equation.

OPEN SOURCE

In a sharp reversal from last year, open source deployments have dropped 25%, to where the trend over four years is now in the negative (see Graph 15). The quality and safety of open source applications are evaluated pretty much the same way as they have been over time, so it seems that the story lies once again in the perceived business value of open source apps.

The number of respondents who cite a lack of business need as a reason for not deploying open source applications has continued to drop sharply, as we see in Graph 16; the logical contrapositive is that respondents perceive a sharp rise in business need to deploy open source software. But equally significant is the fact that the number of respondents who cite a lack of business sponsorship as the reason for not deploying open source applications has more than tripled since 2006.

Similarly to last year, a large majority (85%) of this sponsor-less group works in IT. This is probably due to a combination of economics and communication. As we saw last year, even in an environment of robust IT hiring, open source deployments had a hard time finding business sponsors, implying that businesspeople did not necessarily agree with IT's assessment of the business wisdom of open source. There's no logical reason to assume that this communication divide has gotten any better. On the other hand, the current rate of hiring has dropped by almost half, and the rate of downsizing has more than doubled compared to last year (see Graph 1). The outlook for the coming year was a bit less downward at the time our survey was distributed, but I suspect that at this writing the 2009 hiring forecast is pretty gloomy. So given the preexisting problem of alignment between business and IT as to the business need for deploying open source applications, it stands to reason that those holding the purse strings would have no problem sweeping the open source question right off the table when it comes to decisions about how to spend money. Even though one could argue that proprietary solutions cost more money, in the previous article in this issue, Dennis Adams notes that budgetary decision makers have not been shown that open source is a viable alternative.

IT STAFFING AND OUTSOURCING

Staffing has taken an expected hit. While still about half of companies thankfully remain in a stable IT hiring situation, only a quarter of companies are currently hiring and the remaining 25% are downsizing (see Graph 1). In parallel with this movement, Graph 5 shows that project management as a hiring category has dropped by nearly half, indicating that there are simply fewer IT projects being undertaken. Outsourcing has continued an upward trend, with overall 22% more companies outsourcing now as compared to 2007 (see Graph 3). Companies continue to use nonemployee personnel to provide needed skills while preserving the ability to quickly shed staff as a cost-savings strategy.

Those companies still hiring are doing so with a more narrow focus, not hiring so many different kinds of skills at once, as evidenced by the drop in hiring across all categories (see Graph 5). This means that the overall number of jobs available is down even more than the data here would suggest.

Notably, the mix of IT skills being hired reveals that strategic thinking has not fled before the financial storm. In particular, enterprise architecture and business analysis have now become the number two and three skill areas being hired. About one-third of companies are hiring enterprise architects and business analysts, which is now quite close to the 40% that are hiring application developers, still the top hiring category. While companies are responding to economic conditions with quick tactical maneuvers such as controlling staff levels, they are at the same time continuing to invest in strategic pursuits. Such strategic IT investments have the advantage of contributing both to creating efficiencies over the longer term as well as to supporting value creation via greater nimbleness. A relatively greater emphasis on understanding business need and on properly positioning IT investments to apprehend those needs in the large are wise choices for companies that are seeking both agility and sustainability. A near-term need to cut back development activity does not reduce the need to keep a focus on the big picture.

The skill areas being outsourced are down across the board, as seen in Graph 18, with the notable exception of enterprise architecture, which has risen by one-third since 2007, although as a whole EA is outsourced by only a small 8% of 2009 respondents. I have scratched my head a bit over the notion of outsourcing EA, because in the long run I just don't see how outsourcing strategic work like enterprise architecture can be sustainable.¹ However, a look in my own mirror reflects that I myself have been involved for nearly two years with a client in establishing and building out an EA program and a suite of reusable EA assets, and the client still has a ways to go before becoming self-sufficient.

The upshot is that, as with any endeavor new to a company, companies increasingly have to look outside to bring in the necessary EA expertise. Indeed, the number of companies with EA programs has continued a steady upward trend to nearly 70% (see Graph 11).

WEB SERVICES AND SOA

The use of Web services has not slowed down, as 90% of 2009 respondents report that they use Web services in some fashion (see Graph 19). The use of Web services for local application-specific purposes still predominates at more than 50%. I won't get on my soapbox about that one again this time around (you can read my complaints in the 2007 and 2008 trends issues of CBR²). Rather, I take heart in the fact that other more "cross-border" uses continue to grow. For example, 30% of companies report having deployed common business services across the enterprise, and services are used for point-to-point connections between applications by more than 40%.

SOA programs are now in place in more than 60% of companies, a 13% increase over the past two years (see Graph 20). The shift in focus of SOA shows how companies are using this strategic architectural approach to respond to the economic downturn. Indeed the impetus behind SOA programs has changed dramatically: Graph 21 shows that retiring legacy technology jumped 53% since last year and nearly 83% over the past two years as a primary driver for SOA, while the goal of reducing IT costs rose by nearly 20%.

There are clear economic and risk mitigation incentives for retiring legacy technology. Most mainframe systems are not incrementally scalable; several of my clients are faced with the choice of either dropping many millions of dollars on "another mainframe" or moving into "modern," distributed computing. And there are other pressures: every year the technology is a year older, people with the skills to maintain and advance it are a year closer to retirement, and another crop of new IT people have come up through the ranks without the required skills or the motivation to acquire those "legacy" skills.

Many companies have adopted a mainframe migration strategy of "peeling the onion." First some reporting and analysis functions are ported to Web applications backed by data warehouses and data marts fed by data replication technologies. Then things like customer inquiry, product setup, and some lighter-duty transactional functions are ported. Eventually what will be left on the mainframe are the core business processes that execute lots of embedded business rules or process millions of transactions a day or both.

SOA is a natural approach for effecting this kind of strategy. As each business function is identified, the mainframe code is carefully inspected to ascertain the boundaries of the business function implementation; the business functionality is ported to a new application hosted in a server environment; and a set of services is built to interface between the two. By their nature, the kinds of functions first ported are low-risk, low-volume, and nonmission-critical. Over time, as teams get better at building service-enabled systems, they attack the modules of the mainframe that are more and more central to core business operations. In this way, the use of services/ SOA is allowed to mature in step with the criticality of the business functionality it is used to implement.

On the value creation side of the coin, we see in Graph 21 that the imperative to exploit strategic competitive opportunities has declined by a quarter over the past two years as a key driver for SOA.

The focus for SOA is clearly on saving money rather than making money, to put it in Dennis's simple terms. The economic landscape is doubtless a big factor. But there's another part to this, which is that as more and more companies gain experience in SOA, it has become obvious that the story painted by vendors is much harder to achieve than they make it out to be. (If this comes as a surprise to you, please e-mail me from your parallel universe!) The notion that SOA enables you to respond to volatile market conditions by quickly assembling agile applications out of reusable large-grained business services is some sort of software development nirvana; indeed, it's just a variation on the same theme that vendors have used for years to push technology solutions to business problems. On the other hand, as companies mature in their employment of SOA, they are finding that it is a viable means for apprehending both bottom-up concerns such as integration as well as top-down needs such as factoring out common business functionality into shared business services.

SOA Sponsorship and Impediments

While SOA is increasingly finding its place at the IT table, the story is different at the houses of our business brethren, who are still cautious about giving SOA a salad fork.

The trend of CIOs being the top-level SOA sponsor continues dramatically upward, now approaching half of all SOA programs. One could argue that this makes sense, given the decided shift toward IT cost savings as the primary driver for SOA, as discussed above. After all, who needs business buy-in or sponsorship if what you're trying to do is cut IT costs? (Sirens start blaring here.) Lest we forget, IT is a business investment for the purpose of supporting, enabling, and (hopefully) furthering the cause of business. Every IT dollar you cut will have an effect on things that businesspeople care about and depend on; it's not just a concern for techies. Insufficient sponsorship is now the leading obstacle to the success of SOA programs (see Graph 10).

Indeed, the number of companies experiencing significant impediments to instituting SOA has gone way up, from just over half of companies in 2007 to nearly two-thirds today, as evidenced by Graph 10. And as before, the problems being experienced are all related to how well SOA is meeting business demands: according to the 2009 results, insufficient sponsorship and lack of business buy-in are being experienced by some 30% of companies, with failure to show visible ROI coming in close behind at 26%.

So while investment in SOA is growing, business enthusiasm is waning. I agree with Dennis that SOA implementations that don't meet their business goals but rather create business disruptions will face the axe. If we don't fix this, pretty soon the graphs are going to cross, and we're going to have business executives sounding like Desi Arnaz, as the frazzled TV husband in I Love Lucy, telling CIOs that "you got some `splainin' to do!"

Having said all of the foregoing, it is remarkable that the number one business driver for SOA cited by respondents is still for IT to be more responsive to the business (see Graph 21). Unfortunately, we've seen that the pervasive IT centricity of SOA programs is causing them to struggle for success in this regard. This is notwithstanding the shift away from the expectation of SOA as being a strategic value creation technology. In short, the economic picture doesn't change the fact that there's still the expectation for business-responsive IT. So there will have to be some balance between top-down, business-driven and bottom-up, technology- driven SOA. There's no harm in abandoning the pie- in-the-sky vendor story, but my point from last year's CBR still stands: there needs to be a better partnership between business and IT in order for SOA to be successful.

ENTERPRISE ARCHITECTURE

As further evidence that companies continue to invest in strategic initiatives even under harsh economic conditions, we see in Graph 11 that nearly 70% of respondents now have an EA program in place, representing a 25% increase in the space of two years.

Primary Focus of EA

In looking at what EA programs are primarily focusing on, it seems that money-saving objectives are the clear winners, particularly those with a relatively short time horizon for realizing return.

Implementing technology standards is a basic and time-honored tool for managing IT costs. Reducing the diversity of the technology portfolio limits the number of different platforms that need to be supported as well as the skill base that has to be retained, and it also enables economies of scale in terms of purchasing technology products from vendors.

It is thus not surprising that the implementation of technology and/or data standards has shot back up this year some 30% as the primary focus of EA, after a big drop the previous year, as shown in Graph 13. In a sense, this represents a return to the basics of, as Dennis says, rationalizing IT architecture: managing technology diversity is the most obvious and easily measurable objectives of EA efforts and thus is usually the first to be undertaken and never realistically goes away.

Among the other cost-saving objectives for EA, a steady 25% (approximately) of companies continue to work toward common application architecture with their EA programs, while application integration dropped a precipitous 55%, following a big increase the year before.

Instituting common application architecture is another money-saving measure via standardization in that it makes applications easier, and thus cheaper, to develop and maintain. However, useful application architecture standards take time and money to develop, let alone to put in place and train people on. So the time horizon for ROI is relatively long. However, much of the investment in standardized application architecture is front-loaded: once some useful application patterns have been formalized and developers are familiar with them, it's just as cost-effective in the immediate term to use them versus not to.

Application integration can contribute both to saving money and to making money. On the one hand, it represents a kind of standardization that is yet another level removed in terms of its ability to effect cost savings because it requires some level of application architecture and technology standardization to be in place and its payback is harder to measure.

On the other hand, the business effect of better application integration is an improved customer experience and enhanced business agility.

EA Drivers and Sponsorship

The imperative to reduce IT cost has fallen to second place this year as a driver for EA, with the number one driver being the formal development of enterprise-level business requirements, which has jumped nearly 40% over two years (see Graph 12). This is seemingly in contrast to the focus on cost-saving objectives. However, a closer look at the data provides some further insight.

Of those respondents who do report IT cost reduction as an EA imperative, nearly 70% have either technology/ data standards or common application architecture as their primary EA focus. On the other hand, where key drivers of requirements for EA are formal development of enterprise business requirements or alignment to a defined enterprise business architecture, the primary EA focus for 56% of respondents is still on cost-saving objectives, namely implementing technology and/or data standards or on common application architecture. But 26% of them also report business process integration as a key focus.

The growth in business outcomes as key drivers for EA requirements is welcome news, for the reason that IT-centric EA programs frequently run into problems, as I discussed in the 2008 trends issue of CBR. The growing business emphasis is also supported by the fact that business units are starting to take increased ownership of requirements for EA. Indeed, while the typical ownership by a central IT group has fallen some 14% over two years, Graph 22 indicates that businesspeople have stepped up to the plate in large fashion in 2009, with nearly a quarter of respondents reporting that business units have primary ownership of EA requirements and another 20% going to IT organizations aligned with lines of business.

EMERGING TRENDS

This is the second year that we have had survey questions about some emerging technologies. While it is difficult to detect much in the way of trends, given just two data points, there are a couple of things that are worthwhile to note.

Enterprise 2.0

First of all, it occurs to me that the question about using Web 2.0 technologies is really a subset of the overall Enterprise 2.0 umbrella. If we accept Andrew McAfee's definition of Enterprise 2.0 as "the use of emergent social software platforms within companies, or between companies and their partners or customers," then Web 2.0 is simply getting a little farther into the technology aspect.³

Looking at these questions together, then, three things stand out in Graph 9. One is that between 25% and 30% of companies are now engaged in early experimentation with Enterprise/Web 2.0, although still around a third haven't taken any steps thus far. The other piece is that companies are using social software to reach out to customers in much greater numbers, in a big increase to about 20% of respondents this year.

For Enterprise 2.0 to really be successful, companies will have to rethink their relationship with their customers, their markets, and their employees, and they'll have to look at how information serves these relationships in a completely new light. Exactly what that looks like I can't say, but there is ample evidence that this will require much more than a new face on the old business.

The exploding world of online social environments functions mostly as a means for people to engage around their passions or to provide diversions, titillations, or pseudoaddictive e-potions for überconnected youth. I know that sounds like a polemic, but my point is that the notion of Enterprise 2.0 has some built-in impediments to it. The enterprise imperative of profitable commerce doesn't translate easily into the type of group-minded information sharing that exists in the social networking world. The one commercial model that has taken deep roots in this environment is advertising. Although banner ads are now being rapidly replaced with much more sophisticated targeting, it's still all about pushing stuff at people when they're trying to do something else. The money-making proposition is pretty simple: either you provide ad-pushing services and get paid for it, or you pay to get your ads pushed and hope that they translate into sales. It's pretty easy to measure ad conversion if you're selling relatively low-cost things to retail customers. But for the rest of the business world, the translation to a new kind of relationship with customers is considerably murkier.

That said, it stands to reason that interaction with customers would be an early avenue for companies to foray into Enterprise 2.0. It's almost becoming an expectation that a company should have a corporate blog if it is to demonstrate any sort of cachet. While there are examples of internal corporate blogs, the success stories seem to lie in the blogs that are able to engage customers in a conversation, not necessarily about their products or services but rather about the goings-on, the news tidbits, the insider view, and the cultural aspect of companies. So there's a movement toward engaging customers in new ways; in most cases it's not part of some clearly articulated strategy, but rather is an attempt to try out the new stuff to see what happens.

The use of Enterprise 2.0 as a transformational force for a company's relationship to its employees has even much greater obstacles. The vast majority of people working within any given business enterprise are doing so because it's their job; people are not there doing what they're doing because they are primarily motivated by passion. People interact in online social networking when they go home. To the degree that they do it at work, it's probably supposed to be during their breaks; more likely it's not supposed to be happening in the workplace at all.

Virtual Worlds

There are fewer companies looking into virtual worlds in the 2009 survey than last year, although notably 5% of respondents report that they have already launched implementations. I don't expect the virtual world phenomenon to take off in relationship to business enterprises as long as it is shaped along the lines of Linden Labs's Second Life, which is currently the 800-pound gorilla. About a year ago, I had the opportunity to talk with an engineer from Linden Labs. I asked him what possible benefit a company could derive from having a presence in Second Life. He gave me the example of conducting virtual meetings; to me, that is the perfect counterexample.

The reason is that such a virtual environment is founded on lies, the most profound being the avatars by which people represent themselves to the "world." This is a powerful and very deceptive visual layer on top of the characteristics of online social environments that trumps all of the other sensory cues. Anybody who has spent any time at all in Second Life knows that all of the avatars are fantastically oversized and oversexed, not to mention that many of the "females" are virtual stand-ins for users that are in fact men. This is then painted over with fantastic-sounding fake names, and everyone is then free to run around in a world where nothing is real, and in fact, nobody wants it to be real.

So now consider the virtual meeting. Of course, there are secure subenvironments and methods for using real names and so on. But any of the already mature online collaboration environments that combine video, voice, and document/workspace sharing are already vastly superior to what a virtual world meeting can offer. What's more, Second Life has generated a huge set of expectations that must now be overcome in order for anything meaningful to occur within its confines from a business enterprise perspective. I fail to see how a world in which every relationship is by nature completely suspect can translate into a viable means for effecting any meaningful aspect of commerce for the vast majority of companies out there. Unless you're selling bodily accessories in the virtual world itself, don't waste your time.

A Steady Hand in Turbulent Times

The recent downward economic spiral has predictably put IT in the spotlight as a target for cost reduction. As I've noted in past years, the fact that IT has historically been viewed as an expense while at the same bearing the expectation of delivering strategic value has caused a long-lasting macro-headache that still plagues corporate IT across all industries.

Despite the increased pressure to save money, IT organizations are not just running for the tall grass. Combined with rather predictable cost-saving measures, we see solid trends toward maintaining strategic focus over the long haul. Further growth in EA programs, development of SOA, and continued movement into the nascent Enterprise 2.0 arena are strategic ventures with a relatively long-term payback that deserve sustained investment and organizational commitment.

On the expense side of the equation, headcount and technology purchases are the most obvious places to cut costs because the arithmetic is easy. Time-honored tactical hiring and outsourcing realignments are of course in order. However, this should be counterbalanced with a steady eye on the big picture. For example, strategic hiring for enterprise architects and business analysts can't be quickly ramped down and then back up again because of the relatively long break-in period for these skill areas.

Open source technology is increasingly vulnerable under the current economic conditions. Although IT people may see open source as strategic, the business value creation proposition has not been adequately established. Corporate IT shops wanting to move forward with open source are going to have to come up with a different story. Unless you're a startup that hasn't already made significant investment in vendor platforms, it will be hard to position open source as strategic absent some clear numbers to back up the cost-cutting angle.

Investment in Web services and SOA should focus on pragmatic growth of competency by focusing on application decoupling and enterprise integration. Forget the vendor hype: the top-down story of assembling applications out of reusable services will take years to develop into anything meaningful. What's more, it will require you to be really good at both top-down analysis and program management and governance, all of which take time and hard knocks to develop.

Instead, gain some experience with SOA by focusing on legacy retirement and application integration. The services you build will require that you do sufficient top-down analysis (use those business analysts and enterprise architects!) to create a set of useful services that appropriately abstract out those stable business functions that your legacy systems have been running for all these years.

Enterprise architecture is concerned with how to coordinate the pieces of an enterprise into a coherent whole. We create enterprises in order to realize a higher value than that which could be produced by the individual parts of the enterprise. So the very fact that you have an enterprise means that you're trying to get the pieces of the enterprise to work together in support of the enterprise value proposition. Therefore, every enterprise has EA: the question is the degree to which you are conscious of it. EA is not really an option, but there are lots of options for how to go about it. Clearly cost containment will always be a valid aim of EA (e.g., standardizing technology, common application architecture). But EA is about the "planful" approach to realizing enterprise value with IT -- as such it should not be subject to tactical financial machinations.

Emerging technologies deserve some attention, within reason. Focus your Enterprise 2.0 efforts on online social environments that engage customers in an indirect, nonsales manner to build community awareness and brand loyalty. Such investment is the most obvious and attainable way to get going with Enterprise 2.0 because blogs and other social media have demonstrated their ability to get people connected around their passions and interests. Figuring out how to engage people around those passions and interests in a way that brings them closer to your company and its offerings is the key.

In summary, tactical maneuvers are the order of the day in IT organizations across the board. However, those companies that sustain their strategic investments while making the necessary hard cuts will realize a greater payback.

ENDNOTES

¹ van Tyn, Jeroen. "EA Means Signing Up for Change." Cutter Consortium Enterprise Architecture E-Mail Advisor, 13 August 2008.

² Piccoli, Gabriele. "Starting Off the New Year Looking by Looking Back." Cutter Benchmark Review, Vol. 8, No. 1, 2008; Piccoli, Gabriele. "Trends for 2007: Looking Back to Look Ahead." Cutter Benchmark Review, Vol. 7, No. 1, 2007.

³ McAfee, Andrew. "Enterprise 2.0, Version 2.0." AndrewMcafee.org, 27 May 2006 (http://andrewmcafee.org/blog/?m=200605).

ABOUT THE AUTHOR