GLG News by Paul Massie

Implications: Red Hat’s acquisition of Qumranet will strengthen their virtualization offerings, thus providing more competition for VMWare and Microsoft.  Qumranet’s technology will move Red Hat significantly forward, particularly on desktop virtualization.  This is starting to raise the question of whether virtualization will be able to survive long-term as an independent product.

Analysis:  VMWare’s long-time domination of the virtualization market is increasingly under attack.  Microsoft’s virtualization efforts are well publicized, although they do not yet provide the full feature set that VMWare offers.  Citrix’s XEN-based products are very close to VMWare in terms of features, and the next release of XEN suggests they will close the gap further in the next year or so.  Oracle’s efforts, also XEN-based, have received some publicity but appear somewhat lacking in substance.  Still, one can never count Oracle out!  Red Hat and Novell’s SuSE have been offering some virtualization in the later releases.  This is also XEN-based and currently somewhat limited in scope and functionality.

Red Hat’s purchase of Qumranet signals their intent to move aggressively into the overall virtualization market.  Red Hat has an impressive installed base in servers, especially when the actual number of installations is considered.  Since many of the installations of Red Hat are never licensed it’s not possible to ascertain the actual installed base, but based on the revenue figures Red Hat is showing, combined with anecdotal accounts of the relatively low percentage of Red Hat installations actually licensed, suggests strongly the actual installed base is quite significant.  If/when Red Hat can provide compelling virtualization technology as part of their standard offering, it is reasonable to expect a relatively high adoption rate.

Desktop virtualization is a technology that has received a very lukewarm reception.  While a few enterprises have successfully rolled out desktop virtualization programs, they are clearly a small minority.  Qumranet’s SolidICE technology provides an elegant way of doing desktop virtualization, and when combined with Red Hat running on servers provides a very economical way to implement desktop virtualization. 

Desktop virtualization today has many challenges.  One of this biggest is the cost.  While touted as a cost-saving program, actual costs savings can be surprisingly difficult to realize once the various costs of the servers, the virtualization software and the client platform are considered.  As the cost of the virtualization software gets lower the economics get better and the justification becomes easier.  A Red Hat platform with elegant desktop virtualization becomes a much more attractive alternative and thus could gain significant traction.

All of this starts to broach the question of whether virtualization software has a long-term future as an independent product.  As Microsoft and Red Hat offer virtualization software as part of their operating systems, the business proposition of virtualization as a standalone product starts to look shaky.  This certainly is not something that will happen quickly, since both Microsoft’s and Red Hat’s offerings still have a long way to go.  Looking 3-5 years into the future, however, one must ask whether VMWare and Citrix virtualization products will be commanding a large percentage of the market, or will they be niche players?  Will these products be relegated to high-end, specialized installations, while the vast majority of the virtualization market is controlled by the operating system vendors?  Even if this is the case, such a niche market promises to be quite large, but not perhaps what VMWare’s investors are hoping for!  

Implications: Both the server and PC markets seem to have held up very well during the past months, and barring a catastrophic collapse of the economy, likely will continue to do well.  Credit must go to the innovations in both technology and business processes achieved primarily by HP, Dell and Intel.

Analysis: The ongoing economic problems do not appear to be having much impact on computer hardware sales, whether servers or PCs.  Numbers just released show strong performances in particular from HP and Dell, with IBM also doing well.  Sun Microsystems continues to slump, but that is to be expected given the way x86 servers continue to eat into the Sparc/Solaris market.

It is important to understand the reasons behind these numbers if one is to feel secure about the future of the server and PC markets.  The first factor is that the companies doing well – HP and Dell – are large, multi-national firms with extensive sales outside the US, so as long as the economic troubles are focused on the US their overall sales numbers are buffered by sales to the rest of the world.  This also implies that any companies selling only in the US will suffer by comparison, but most of those have already been acquired, so this is not much of a factor. 

Ironically, today’s success seems to be rooted in yesterday’s troubles!  Three or four years ago most of these companies were having some level of difficulties.  HP’s PC division was a perennial money-loser in the pre-Hurd days.  Dell was in serious difficulties before Michael Dell re-assumed control.  Intel was being upstaged by AMD.  All three companies made concerted efforts to improve every aspect of their business, from technology through quality and customer service.  Intel leapfrogged AMD with chips released in 2006, HP re-energized both their PC and x86 server divisions with new designs and efficiencies, and Dell improved quality and customer service while expanding their markets.

The timing could not have been better for all three.  Had their respective problems occurred at the same time as the economic slowdown the impact would have been far greater.  As it was, by the time the economy had slowed all three companies had already gained momentum and were poised for success.

The root of much of today’s success for these companies lies with Intel’s processor technology.  The processors driving this hardware have improved so much in the past two years that the servers, desktops and laptops utilizing those processors make a compelling argument for upgrading.  The cost/price ratio of new servers is so much better than servers 3-5 years old that enterprises continue to buy even when money is limited.  The ROI for these efficient new servers is sufficiently great that new purchases are easy to justify.

In the PC market the new machines are relatively inexpensive compared to 3-4 years ago, and at the same time offer a functionality increase similar to servers.  The applications used on many PCs, whether for business or consumers, have expanded sufficiently so that older machines perform poorly.  At the same time, pricing on new machines is aggressive enough to encourage upgrading.

The big question, of course, is whether today’s happy state will continue?  It appears most likely that it will, at least for another year or two.  Despite strong server sales, enterprises still have a large backlog of older and soon-to-be obsolete servers which will need to be replaced.  Consumers and businesses also still have a lot of older machines that need to be replaced.

It may be a different story when the current generation of machines has finally saturated the market, most likely in 18-24 months.  At that time everyone will have a server or PC powerful enough to handle the workload, so unless some new and compelling driver appears the market may well slow.  One such driver could be the end of the current economic troubles.  If that occurs in one or two years it could be a sufficient driver to give the hardware market legs for another year or two.

Implications: The next evolution of Xen software puts further pressure on VMWare’s market share and margins.  This is a clear indication that VMWare’s business model will require a major overhaul sometime in the next year or two.  Microsoft will be less impacted since they have less today. 

Analysis: From the technology point of view there are only three important competitors in virtualization, VMWare, Microsoft and Xen.  Many people have probably never heard of Xen, although XenSource (bought by Citrix last year) is better known.  Xen is the open source software that lies at the heart of the products XenSource was/is producing.  It also provides the technology for a number of other virtualization businesses, such as Citrix, Virtual Iron, Red Hat, SUSE, and Oracle.  It has been widely accepted by all the major server vendors.

The key concept here is that whenever the open-source Xen software incorporates new functionality, that same functionality will appear in a number of commercial products from Citrix and others sometime during the following months.  The companies that are really focused on virtualization (e.g., Citrix and Virtual Iron) will attempt to leverage that new functionality into compelling products, while other companies that include virtualization as an add-on (Red Hat and SuSE) will tend to pass it along more or less as is.  In either case, new Xen functions implies a wide array of products competing with VMWare.

VMWare today has a significant lead over both Xen and Microsoft in terms of technology and they use that lead to justify their pricing structure.  Both Microsoft and Xen are attempting to reduce that lead, although the Xen family of companies is ahead of Microsoft.  If Microsoft and the Xen family can eliminate most of VMWare’s technology lead, VMWare stands to lose much of their market, since the pricing for both Microsoft and the Xen family of products is significantly lower than VMWare’s pricing.

Even if VMWare is able to maintain a technology lead, there will come a point where that will matter less and less.  Once the Microsoft and Xen family of products incorporate all the functions the average IT shop needs any extra functionality VMWare offers will matter little.  That time is probably within the next 12-18 months, after which few CIOs will be willing to pay VMWare’s premium for additional functionality they don’t really need.

Open source has not (yet) been particularly successful in enterprise applications, but has done very well in certain infrastructure areas.  Linux has decimated the Unix market.  MySQL has not seriously damaged Oracle or IBM’s DB2, but it has gained enough market share to get their attention.  The virtualization market seems poised to go the same direction as Linux. 

Linux grew so fast because the combination of Linux and x86 servers was far more cost-effective than Solaris and Sparc servers or AIX and PowerPC servers.  Xen-based virtualization tools seem poised to have the same impact on VMWare as Linux has had on proprietary Unix.  Microsoft’s virtualization efforts won’t help VMWare, since Microsoft is also providing the core virtualization technology at little or no charge. 

Xen 3.3 incorporates functionality that starts to provide most of what IT shops need.  If not this release, then the next one or two releases will have “enough” technology for IT, at which point VMWare will only be able to compete by lowering prices dramatically.  Even then, it’s going to be an uphill battle for them.

Implications: Intel’s next-gen Nehalem processor ups the performance bar, making it increasingly difficult for AMD.  Server manufacturers and software vendors will also be impacted, positively or negatively, by this generation of processors, which will also provide further impetus to cloud computing.

Analysis: AMD has held an architectural advantage over Intel for the last several years, although Intel has been able to regain the processor performance crown despite these inherent limitations.  Intel’s method of accessing memory, the Front-Side Bus (FSB), is an archaic approach that has limited processor performance.  In addition, Intel’s quad-core offerings have been more single-core processors tied together than designed-from-the-ground-up quad-core processors.  AMD has offered an integrated memory controller and multiple cores on the same die for some time.

Intel’s Nehalem processor removes these barriers.  It is due out very soon on the desktop, followed by servers and mobile during the next few months.  For the first time, Intel is offering a processor with an integrated memory controller.  Nehalem will also offer four cores on the same die, with the Hyper-Threading (remember the Pentium 4 and Hyper-Threading?) architecture revived.  This means Nehalem will offer top performance for four application “threads” simultaneously, and possibly as many as eight threads under some circumstances.  Since Intel’s processor performance was already superior to AMD’s, this does not bode well for AMD’s future prospects.

The greatest impact of this architectural change, however, will be in the server market.  Desktop computers are rarely able to fully utilize four cores today, and with the new and improved Nehalem it becomes even less likely.  The actual performance lead Intel has thus becomes somewhat academic for most desktop computers, gaming being a notable exception.  Servers, by contrast, are able to fully utilize multiple cores.  They also tend to run very large applications that require more memory and thus the integrated memory controller is more of a factor.  This is a big reason why AMD has been able to maintain reasonable market share in the server market.  Both the integrated memory controller and Hyper-Threading will likely have much more benefit for servers, thus hitting AMD in one of their remaining strengths.

The full implications go well beyond AMD.  These processors will provide increasingly powerful small servers.  This is likely to impact the large server market even more, since many of the applications that previously required larger servers can now be handled by low-end, commodity servers.  Companies that derive much of their revenue from the low-end server market (Dell, HP) stand to benefit at the expense of those who focus more on high-end servers (Sun Microsystems). 

Applications vendors will also be affected.  Virtualization becomes ever more important as small servers become more powerful.  VMWare will benefit, but so will Microsoft, Citrix and even Virtual Iron.  With Nehalem, an amazing amount of computing power becomes available in a rack of blade servers.  This cries out for virtualization and is going to be a huge driver in the development of the “private cloud” computing model, where enterprises use virtualization tools to create their own internal compute cloud.

Implications: Rumors abound that Cisco might purchase EMC, but that seems very unlikely given the cost involved and the lack of apparent synergy.  Cisco could, however, purchase VMWare.  Cisco has made virtualization a key strategic focus, and adding VMWare would instantly catapult them to the top of the virtualization rankings.

Analysis: As the referenced article states, the probability of Cisco buying EMC seems very low.  The market cap of EMC is currently in excess of $30B, which means the price tag for an acquisition would have to be above $40B.  Cisco has a lot of money, but not that much!  They could probably leverage such a deal and make it happen if the payoff was clear, but in this case the payoff is anything but clear.  EMC is focused on a market that is significantly divergent from Cisco’s strategic focus, so such an acquisition would make little sense for Cisco even if the price were low.  Given that Cisco would have to leverage the house to make the deal work, the probability of such a deal seems close to zero.

VMWare, however, is a very different picture.  VMWare’s current market cap is around $12-15B, so an acquisition could conceivably be done for no more than $18-20B.  It’s still a lot of money, but more within Cisco’s grasp.  It’s also a direction that fits well with Cisco’s market strategy.  Cisco has identified virtualization as one of the key focus areas for their further growth.  Their “VFrame” product line seems to be doing well in the market, but nonetheless Cisco is not a name that pops readily to mind when top virtualization vendors are being discussed. 

Cisco is above all a marketing machine, and nobody knows better than they the importance of mindshare.  They would likely be willing to make a major investment to gain more mindshare in a market they are targeting.  Buying VMWare would achieve that goal instantly, since despite VMWare’s recent troubles they are still the virtualization leader in technology, revenue and mindshare.  Their image may be slightly tarnished recently, but as yet the damage is minor.  Historically Cisco has intimate knowledge of VMWare, since they invested $150M in July of 2007 to buy a small percentage of the company, a percentage they still hold.

EMC still has virtually complete control over VMWare, so it would be up to them to decide whether Cisco could buy all or part of VMWare.  Despite VMWare’s recent drop (collapse?) in market cap there seems little reason to believe EMC has soured on their investment.  EMC doesn’t need the money and virtualization is still a key part of their future plans.  It’s very difficult to picture EMC selling their investment, or even a majority stake, to anyone at this point.

The most likely scenario seems to be Cisco increasing their stake in VMWare without actually purchasing a controlling interest.  Cisco could invest anywhere between $1B and $5B into VMWare without impacting their cash position.  This would give them increased control over VMWare’s future direction and ensure their shared vision would continue to flourish.  They might also consider it a good investment at this point, given VMWare’s recent drop in market cap.

It seems VERY unlikely Cisco will even consider buying EMC.  It also seems unlikely EMC will give up their controlling interest in VMWare.  It does seem reasonable that Cisco could increase their stake in VMWare, which would probably prop up VMWare’s market cap somewhat.  That alone would make the idea attractive to both Cisco and EMC.

Implications: Microsoft’s acquisition of DATAllegro should put to rest any skepticism on whether Microsoft is serious about the data warehouse and Business Intelligence (BI) market.  They have been progressively moving more into BI and data warehouses for several years, but have focused on the low-end BI user.  Now they’re going after the high-end market as well via an appliance. 

Analysis: A number of factors are converging to move BI into the mainstream.  Not least is the fact the ERP market has consolidated and matured, so Oracle and SAP are looking hungrily for new worlds to conquer.  This is shown by their recent acquisition of Hyperion and Business Objects.  With IBM’s acquisition of Cognos, the BI market is clearly consolidating.  At the same time the amount of data enterprises need to manage has exploded and computational power has grown to the point where extracting useful information from that data is feasible.  BI tools provide a presentation layer which is only as good as the data warehouse under it.  No BI system can provide “intelligence” without a data warehouse capable of managing huge amounts of data.

Microsoft is looking for new markets for growth as they are increasingly confronted by the law of large numbers.  BI is a market they have targeted as key to their future growth.  The last release of their database, SQL Server 2005, was a serious contender in the data warehouse arena, albeit still generally for smaller enterprises.  The current release, SQL Server 2008, further enhances scalability and capability for data warehouses. However, it still does not offer the performance required for a really high-end data warehouse, which could have upwards of 100 TB of data.  Only a hardware appliance can handle that amount of data.

The high-end data warehouse market is not particularly large in terms of numbers, but the revenue is quite significant.  The market leader today, Teradata, has annual revenue approaching $2B.  HP is also focusing on the high-end data warehouse market.  DATAllegro has specifically targeted Teradata with their offerings in the recent past.  Their technology is generally considered comparable or better than Teradata’s, although their revenues are much smaller. 

Microsoft’s acquisition of DATAllegro clearly spells trouble for Teradata, Oracle, HP and any other would-be players in this market.  More importantly, it represents a key building block in Microsoft’s BI product portfolio.  They already have low- to mid-range data warehouse capability with SQL Server.  Now they can address the highest-end data warehouse requirements.  They already have low- to mid-range BI capability.  All they need is a high-end BI offering and they will be the only vendor to address the entire BI product suite from low- to high-end with both data warehouse and BI products.

Implications: HP maintains worldwide leadership, but Dell is staging a comeback.  HP may be able to leverage their wide array of products for an advantage.  The second-tier vendors will continue to do well in places but seem to have little chance of unseating either Dell or HP.  Apple is a dark horse poised to improve in this market.

Analysis:  HP’s resurgence over the last few years is well documented and they are now the clear market leader in worldwide PC sales.  Considering PCs were something of a loss-leader for them 3-4 years ago this represents a remarkable turnaround.  Dell in the interim has had their problems, with quality issues and declining market share.  They began their turnaround some time after HP, but appear to be seeing solid results at this point.  It is telling that Dell was able to take market share from HP in the US during the last quarter.  This sounds more like the Dell of old, rather than the stumbling giant they had become a couple of years ago.  Nonetheless, given HP’s worldwide strength it seems very unlikely that Dell will be a threat to HP on a worldwide basis.  Dell is working to enhance their footprint outside the US, but they have a long way to go to match HP.

The second tier vendors, Lenovo, Acer and Toshiba, lag both HP and Dell by a wide margin in worldwide shipments, although they are all doing well in particular areas.  Lenovo leverages their IBM heritage of engineering and quality for enterprise sales, although they frequently suffer in price comparisons.  Acer does best in Europe, and (particularly in the US) for cost-effective  consumer sales.  Toshiba is still a premium brand but is not a significant player in the US market today.

Bringing up the back is Apple, with a potential that belies their current worldwide sixth place.  Since Apple has essentially ignored enterprise sales for many years it is remarkable they are able to hold a place this high, especially with their focus on US sales.  Apple is benefiting from the current backlash, in the US especially, against Microsoft’s Vista push.  This is driving a number of Wintel purchasers into Apple’s arms, and as XP gets older and harder to obtain over the next year or so, this push is likely to intensify.

Nobody can predict how the total PC market will develop during the next six months or so.  Will the almost-recession bring about a slow-down?  Or will pent-up demand force purchases?  Most companies today are reluctant to replace PCs, yet many of their PCs are getting so old they must be replaced.  Actual corporate earnings are not terrible, so most companies have enough money they “can” replace PCs if they really need to.  There has been economic doom and gloom for long enough now that many of the companies that postponed replacing PCs will soon need to move ahead.  It seems likely that, barring a significant additional downturn in the economy, PC buying will be reasonably healthy during the latter half of 2008.

HP will certainly continue to do well.  They are on a roll and their recent market share loss in the US is probably more attributable to Dell improving rather than HP having problems.  Dell continues to recover from their problems and seems poised to gain more market share.  Apple seems very well positioned to gain 2-3 points of market share during the next 12-18 months, given Vista’s continuing problems and Apple’s heightened profile.  The other three, Lenovo, Acer and Toshiba are likely to continue to inhabit the middle area with little significant change.  No other company seems to have any chance of survival in this market.

Implications: At a market cap of $7-8B and $2+B in the bank, along with healthy free cash flow, Sun is starting to look ripe for an acquisition.  It’s becoming increasingly clear that Sun’s current business model isn’t working, but their IP, customer base, brand recognition and employee base makes them attractive.

Analysis: Time is running out on Sun as they attempt to remain independent and prove their business model is working.  They are still in an enviable position in many respects:  They have more than $2B in cash, a solid free cash flow, huge IP and employee assets, unparalleled brand recognition and a blue-chip customer base.  Despite these factors, their business seems to be stagnating or slowly declining.  They’re essentially treading water in business terms – doing well enough to keep going but not able to demonstrate solid growth.  At the same time their competition is going from strength to strength and pulling away. 

The commodity server market is taking more and more of the overall server market, largely at the expense of the proprietary Unix market (e.g., Sparc).  IBM is maintaining a solid leadership position, HP has staged a remarkable comeback in the last few years and shows no signs of slowing, and Dell seems to be making a solid recovery from their problems of a couple of years ago. 

In the storage market EMC is both moving ahead with their basic product mix while offering a wider array of storage-related software.  Network Appliance and Hitachi are doing well.  Meanwhile Sun, despite their arguably excellent products, seems to be losing mindshare as well as market share.

In the open source world Sun has a major presence, with Open Solaris and MySQL.  Unfortunately, Sun has not yet demonstrated how they plan to monetize open source, at least not at the level of revenue they need.  Red Hat has made a successful business model from open source, but their revenues are still a small fraction of what Sun needs if open source is to rescue them.

At the same time, Sun still has around 30,000 employees and a cost structure rooted in the days when they had a high-margin revenue stream from proprietary servers.  Their Sparc/Solaris-based server market is starting to look like IBM’s mainframe market in the late Eighties or early Nineties – still profitable but an ever-shrinking share of the market.  Sun cannot expect that product line to sustain them.   Commodity (x86) servers are a low-margin business that cannot maintain Sun’s business model.  Open source similarly doesn’t have the margins for their business model. 

These are the factors contributing to Sun’s current market cap, which is likely to go much lower if/when the large cap funds are forced to drop Sun.  At a market cap of $5-8B Sun becomes a prime target for a takeover.  As discussed in the article, Fujitsu must be considered a prime candidate, given the existing partnership and synergies.  HP should also be considered a strong possibility.  Although there is an overlap on some of their products, Sun has sufficient independent resources to be quite attractive to HP at their current market cap.  Further, with HP’s recent experience in cost-cutting, it seems likely that HP could make Sun a very profitable company within a couple of years, albeit perhaps with a headcount closer to 15,000 than 30,000. 

An acquisition by IBM doesn’t seem likely.  The fit just isn’t that good and the product overlap is very strong, both on hardware and software.  Dell is a possibility, but the culture clash would seem to doom that marriage.  The high-cost R&D-focused culture of Sun would not fit with Dell’s low-cost, low-R&D approach.  A private equity buyout could also be in the cards, if the funds are available somewhere in today’s financial climate.  The price is low enough to make such a buy-out reasonable, and 2-3 years private would be time enough to strip a lot of the costs out of Sun and prepare the company for re-emergence as a leaner and more profitable entity.

Sun may remain independent, but with their current market cap so low, it would be easy for a suitor to offer a 50-75% premium.  And as Yahoo’s current travails demonstrate, resisting such an offer would be very difficult.

Implications: VMWare is doing great, so how could they be in trouble?  By having both Microsoft and open source invade their market.  Their easy growth is in the past, and now the question becomes whether they can thrive in a highly competitive environment.

Analysis: VMWare essentially created the x86 virtualization market and has been the dominant (and for a long time, the sole) vendor in the market.  As such they have been able to charge premium prices for their products and use those high margins to develop still more innovative products, as well as other plums such as an “eco-friendly” new headquarters.  Their record-setting IPO in 2007 saw their market cap ascend to astronomical heights.  2008 has not been so kind to them.  Growth rates are slowing, with two revenue forecast reductions already, capped by the recent ouster of their co-founder and long-time CEO, Diane Green.

VMWare remains the undisputed technology leader, but their premium pricing has left them vulnerable and put a damper on their continuing growth.  CIOs are quite willing to buy VMWare products in limited quantities to virtualize selected elements of their infrastructure, but not to move into mass virtualization.  Most IT shops have a few areas where the gains from virtualizing provide an immediate ROI even at the prices VMWare charges today.  Few IT shops can find a realistic ROI for virtualizing large portions of their infrastructure at today’s prices.  Consequently, most CIOs have dabbled a little in virtualization, but are playing a “wait-and-see” game before moving any further.  This is one of the primary reasons why virtualization today has only penetrated 5-10% of the available market.

Microsoft is moving aggressively to attack this market with their tried-and-true approach of providing technology that is “good-enough” at dramatically lower prices.  Perhaps even more dangerous to VMWare, however, are the various open-source efforts that are focused on the virtualization market, some of which are now under the umbrella of major software vendors such as Citrix.  Microsoft will charge commodity prices for technology to win market share, but once they have market share their prices tend to move up.  Open source, by contrast, will never charge premium prices, so the various open-source virtualization products essentially put a ceiling on the prices VMWare will be able to charge.  Many CIOs, of course, are still uncomfortable entrusting their infrastructure to open source products, but the Citrix endorsement will overcome many of these reservations.  For the rest, there’s Microsoft.  Also, the virtualization capabilities in Red Hat and SuSE Linux should not be underestimated.  While these offerings are still quite limited compared to what VMWare has, they do take away a portion of the virtualization market that VMWare would otherwise have.  The net cost of these offerings, even with full licensed support, is still zero since they are included with the Linux license.

With these factors in play it starts to become clear why a CEO change was needed.  If VMWare continues business-as-usual, they will fail.  Not today, not tomorrow, and probably not for several years, but their decline in market share and margins will be steady and inevitable.  If VMWare is to be successful they will have to accept lower prices for their current products, while at the same time continuing to innovate with new products.  Their connection to EMC is a potential advantage allowing them to offer a complete “cloud” product that allows enterprises to virtualize their entire infrastructure, not just servers.  At the same time, however, VMWare must be cautious not to burn any bridges with their other partners such as Network Appliance.

VMWare must walk a tightrope and make fundamental changes to thrive in this environment.  This requires a new strategic direction and likely new blood at the top.  Only time will tell whether Paul Maritz is the right person for the job, but certainly he understands the threat from Microsoft and has the credentials to steer the company to success.  Microsoft’s products are not truly competitive today, but they will be in a few months.  VMWare still has a window of opportunity to capitalize on their lead, but that window is measured in months, not years.

Implications: According to the media and several prominent individuals, SaaS is the future of the software business.  The “old-fashioned” licensed software model is doomed.  Yet the most profitable and biggest software companies have only a limited presence in SaaS.  Does this mean they are out of step with the future, or do they actually know something?

Analysis: Articles appear almost daily touting the benefits of SaaS and proclaiming the end of the licensed software model.  Licensed software is a dinosaur that just hasn’t realized it’s dead!  CIO’s should fully embrace SaaS today or be seen as out of touch with the industry!  It’s easy to find articles talking about how a CIO is saving time and money by using a SaaS vendor.  The market cap and growth rates of SaaS vendors such as Salesforce.com and SuccessFactors are the envy of everyone.  Does this mean that Microsoft, Oracle, SAP and others must fully convert their software to a SaaS model or perish?

Perhaps a note of caution is appropriate before dashing headlong into the SaaS model.  Ellison’s comments in the attached article about the difficulty of making SaaS a profitable business must be considered.  Yes, he does have a bias – Oracle makes a lot of money off the “old-fashioned” licensed software model.  Yet for all his failings, Ellison has never been slow to seize any opportunities to make more profits.  If he truly thought SaaS was the place to be, Oracle would now be fully converted to SaaS.  Ellison has quite a lot of experience with SaaS, both from Oracle’s SaaS offerings and his personal investment in NetSuite.  So it’s probably safe to assume there is at least a kernel of wisdom in what he says.

Nonetheless, even if vendors can’t make a lot of profit off SaaS, can it still dominate software?  Not necessarily!  One crucial factor that seems to be continually overlooked in the press is that companies such as Salesforce.com and SuccessFactors are more than SaaS vendors.  It is true they deliver their products via the SaaS model, but it is also true their products are relatively new and built following the latest Web principles, unlike most traditional software vendors.  Software users (that would be Sales people, HR people, Operations and the like) don’t know or even care how the software is delivered.  The business users care about functionality and ease-of-use.  Most successful SaaS vendors today are growing and doing well because they are delivering the software users need and like to use.  The SaaS model they use may facilitate the sale, but it certainly does not make the sale.  No business user ever decides they want to buy some SaaS software.  They decide they want to buy HR or Sales software, and don’t care about the delivery model.

The decision to go SaaS or not, given otherwise equal software functionality and interface, is purely a decision between IT and Finance.  SaaS requires little or no CapEx and no large upfront cash outlay, so SaaS is very attractive to companies needing to conserve cash.  Also, SaaS software requires little staffing or technical expertise by the customer, so it works best for companies with small IT departments, or where IT departments don’t wish to acquire staff and expertise in a particular software package.

SaaS as a delivery model certainly has a robust future, but it’s difficult to see it conquering the software world.  Software that is modern, capable and easy-to-use will absolutely conquer the software world, whether that software is delivered via SaaS or license.  Today’s SaaS success stories will have more challenges when the large vendors start delivering modern, easy-to-use software, such as Oracle’s Fusion Apps promise to be.

Implications: Motorola’s new camera phone won’t be enough to resuscitate their ailing handset business, but it does show that some innovation still lives within Motorola.  This is not the second coming of the RAZR, but is a good beginning if they can follow with other innovative products.

Analysis: Motorola has been dying (literally) for a follow-on to the RAZR, which almost overnight catapulted them into one of the leaders in the mobile handset market.  Their inability to find another success has led to collapsing revenues, an executive exodus, shareholder unrest and the likelihood they’ll divest (or even just close down) their money-losing handset business.

The Motorola MOTOZINE ZN5 is their new camera phone.  There’s nothing new in camera phones, is there?  Many, if not most, mobile phones come with cameras.  Most of life’s most camera-worthy moments occur with little or no warning.  Professional photographers usually have a high-quality camera within reach, so they can record even the most unexpected events.   For everyone else a camera phone has to suffice.  Granted, the pictures (generally 1.0 or 1.5 megapixel with marginal optics) aren’t of very high quality, but they’re better than no pictures at all.

Now Motorola is taking the position that such compromise is no longer necessary.  A normal mobile handset (the ZN5) can now provide high-quality pictures.  They may not be professional-quality pictures, but 5-megapixel and good optics is as good as most point-and-shoot digital cameras provide today.  In addition, the autofocus, flash and software make this a legitimate camera, not just another camera phone.  Viewed strictly as a digital camera, the MOTOZINE ZN5 has to be considered a competitive entry.  

As a phone, the ZN5 doesn’t appear to be outstanding, although it meets all the standard requirements.  It has all the normal bells and whistles expected in high-end mobile phones today, but nothing that appears particularly distinctive.  Viewed strictly as a mobile phone, the ZN5 is just another good, but not outstanding, entry in the field.

The combination of a good quality point-and-shoot digital camera with a solid mobile handset, however, should win some market for Motorola.  Suddenly photos shot with a camera phone are of high enough quality to be considered legitimate photos and not just spur-of-the-moment snaps.  The average digital camera is no longer needed – the mobile handset is just as good.

This is probably not going to rescue Motorola’s handset business, but it is a solid step forward.  It is likely to gain some market for Motorola and may take some market from other handsets and from digital camera vendors.  It does demonstrate some life is left within Motorola’s handset business unit.  If they can follow this with a few more innovative offerings that business might yet recover, but if they depend solely on the ZN5 it won’t be enough.

Implications: The move from rotating disks to Solid State Disks (SSD) is happening faster than most predicted.  Rotating disk vendors such as Seagate will likely see margins erode as they lose the high-end disk business to SSD.

Analysis: SSD has been around in one form or another for many years.  It was available in the Eighties as a pricey add-on for mainframes and in the Nineties for servers, but it has never achieved more than a niche status due to the cost.  A few months ago the CEO of Seagate went on record stating he wasn’t worried about SSD, since the price was still too high and not dropping fast enough to be a concern.  Now it’s starting to look like he may have been mistaken.

Major enterprise suppliers have made a movement into SSD recently.  HP made the announcement in the attached article, Sun is providing SSD as an option on some of their servers and EMC is making it available with their high-end Symmetrix storage units.  It is true that the price is still quite high, and reliability remains a concern.  Nonetheless, when vendors of the size and credibility of these three make a move the market should pay attention.

The most-cited concern is the price.  Even today, however, for some high-performance environments the price is competitive with rotating disks.  The fastest rotating disks available are quite pricey, as are the various pieces with which they are packaged, yet they frequently do not provide acceptable performance.  To offset this issue many enterprise utilize only a small portion of the available space on their rotating storage, which allows them to achieve higher performance but also drives up the price.  When SSD is compared in this kind of environment their performance is clearly superior and their price not so different.

The major vendors have all addressed the reliability issue through various technical tricks.  EMC, for example, says that their current SSD offering provides reliability comparable to that achieved with high-performance rotating disks.  When companies such as HP and EMC offer a product it is safe to assume they have made it reliable and expect it to sell in reasonable numbers.  They clearly see a market today, albeit perhaps a small one.

The key thought here is the speed with which the transition is occurring.  This is not something happening in the next 3-5 years as many have predicted, but rather within the next 1-2 years.  SSD pricing is coming down rapidly, much faster than rotating disk pricing, and technological advancements are happening faster.  Rotating disks are still improving, particularly at the lower end of the market, but fast improvements at the high end of the market are constrained by the laws of physics.  SSDs are not yet facing such barriers.

Rotating disks are not going away any time soon.  People have been predicting for many years that tape would go away, yet most data centers still have a major investment in tape.  What is going away is the high-end market for high-performance premium rotating disks.  SSD is starting to move into that market.  The impact will likely be felt within months, and certainly within a couple of years.  That will leave a robust market for rotating disks, but more as a commodity.  That transition will hurt companies such as Seagate, which today achieves higher margins on the premium rotating disks.

Implications: Oracle is getting close to delivering on their promises about Fusion.  If the final suite of products can match the early previews, they have achieved a “game-changer” in enterprise applications.

Analysis: At Enterprise 2.0 this week Oracle has a booth showing previews of some of their upcoming Fusion applications.  What’s visible at the booth represents a truly remarkable turnaround for a traditional enterprise applications vendor.  They are showing products that have a simple, intuitive user interface and incorporate numerous Web 2.0 features.  They could have (and many expected they would!) simply adopt a few collaboration tools and claim they were Web 2.0-enabled.  Instead they appear to have really thought about how to take the best concepts of Web 2.0 and incorporate those into enterprise applications.

The Oracle Sales Library tool, for example, allows users to upload PowerPoint presentations, which are then dissected slide-by-slide and made available to users by slide.  Each slide can be tagged and rated by users, so once a few presentations are loaded there will be a library of annotated individual slides that can be quickly reconstructed into new presentations.  Today when someone needs to do a new Sales presentation they have to painstakingly go through multiple existing presentations and do cut-and-paste to extract the specific slides they need, or just re-create the whole presentation from scratch.  The OSL will present a library of individual slides that can be trivially assembled into a new presentation.  This will be a killer app for Sales and Marketing people, who can now create presentations customized for each customer with minimal effort.  OSL brings “mass-customization” to the Sales/Marketing presentation world.

The caveats are many, of course.  This is still a demo and not a released product.  Oracle will naturally be showing only the best features at the show.  Also, the entire Oracle applications suite consists of dozens of modules, of which this is only one.  There’s no way of knowing how innovative the remaining modules will be.  It’s really hard to do Web 2.0 in General Ledger!  It’s clear the entire suite is not going to be released in the next couple of months, and in fact it could easily be a year or more to get most of the modules released.

Nonetheless, despite the caveats, if this is how Oracle is doing enterprise applications their competitors should be worried – very worried!  The obvious candidate for concern is SAP, but there are numerous smaller companies that should worry.  Many of the SaaS vendors, such as Salesforce.com, SuccessFactors and others, have made their business based on supplying intuitive, elegant interfaces to enterprise users.  If Oracle is able to provide better interfaces and more innovative tools, it bodes ill for their future.  Further, unlike an ERP system, areas such as Sales Force Automation or Human Performance Management delivered as a service can be rather easily replaced, so Oracle could potentially recover some of the business that has migrated to the various SaaS vendors in the last few years.  

The Oracle of 2012 may well achieve Larry Ellison’s goal of $50B in revenue through organic growth as well as acquisitions. 

Implications: Riverbed jumped into the WAN optimization market with innovative technology and an effective business plan.  Their success has justified the market, but they face an uphill battle against the big network vendors.  Ongoing innovation gives them some hope, but their long-term viability is at risk.

Analysis: In the early days there was application acceleration, based primarily on optimizing various applications protocols, such as Microsoft’s CIFS.  Network optimization took an entirely different tack, focusing on improving data traffic.  Both provided a small but thriving market for a number of vendors such as Packeteer, Radware and others.  Then Riverbed came along with innovative technology that combined the best of both into one relatively inexpensive box.  It was like Christmas for businesses facing escalating network costs and application delays. 

Placing a couple of Riverbed accelerators at both ends of a corporate WAN meant that 4Mb links were enough in many cases – no need to upgrade that Bangalore office to a 20Mb connection!  Of course, some WAN links were under-utilized already, so further acceleration was unnecessary, but as companies became more decentralized WAN links started coming under increasing pressure.  For businesses with over-stressed WAN links, particularly into less-developed countries, Riverbed promised (and delivered) stunning performance gains.  In some cases the ROI could be as little as 3-4 months.  The cost of the entire Riverbed installation might be no more than increased WAN bandwidth would cost for 3-4 months.  Performance gains of 1-5,000% were achievable in the real world.  LAN performance over the WAN was a fact.

Unfortunately for Riverbed, this rosy picture was evident to the networking giants, most notably Juniper and Cisco.  Both have entered this market with competitive products and pricing within the last 12-18 months.  That allows the compelling argument of one less vendor in the corporate network, potentially less TCO, and possibly at some point more effective implementations as this technology is more tightly integrated into routers and switches.  As one Cisco executive allegedly stated, “…we have more engineers working on this product than Riverbed has employees!”  Whether or not that is true, the message is clear.

Riverbed has responded with continuing innovation, but whether that will be sufficient to sustain them for the long-term is an open question.  Products from Cisco and Juniper are competitive today in terms of price, performance and functionality.  As they integrate those products more tightly into their networking products, the potential (and real) market for Riverbed shrinks.  It’s too early to write off Riverbed, but one must be dubious about their long-term prospects.  Naturally, any other independent vendors attempting to survive in this market are embarked on a suicide mission.

Implications: The release schedule for Windows 7 effectively means most enterprises will go directly from XP to 7.  Windows 7 appears to be Vista done right, with problems addressed and some nice functionality added.  If delivered as promised it should maintain Windows’ position in the market.

Analysis: The problems with Microsoft’s Windows Vista are well documented and appear to provide opportunities for Apple and Linux to make inroads into the enterprise desktop/laptop market.  Although Microsoft has vocally defended Vista both in terms of quality and market share, it appears they are actually paying attention to the feedback.  The original release schedule for Windows 7 was “2010”, which generally means sometime late in the year.  The current dates of RTM in November, 2009 with general availability in January, 2010, while technically meeting the letter of “release in 2010” is 9-12 months earlier than many people had expected.

Microsoft is notorious for slipping schedules on Windows projects, but the external pressure here suggests they will ensure this schedule doesn’t slip.  These dates are important because they are early enough that most enterprises can choose to skip Vista entirely and move from XP to 7.  General availability in January, 2010 means enterprises could start doing rollouts late in 2010, which is about the time XP will finally reach its “sell-by” date for enterprises. 

Enterprises usually want to wait until the SP1 version before rolling out Windows, however if the changes are minimal this may not be necessary.  Early information on 7 indicates it is based on the Windows Server 2008 kernel, which has generally been well received.  The hardware requirements for 7 are intended to be comparable to Vista, which will eliminate any hardware upgrade requirements.  Finally, 7 is intended to generally maintain compatibility with Vista software and drivers, which again eliminates a major delay for rollouts.

All of this suggests that 7 can be considered as Vista done right!  In many ways it is starting to look like the SP2 release of XP, which was the time when adoption of XP really took off.  Although Microsoft has been very defensive about Vista, it’s becoming clear that the Windows 7 development team has been listening to the complaints about Vista.

The touch-screen capability, as well as any other new features that may still appear, provide a nice benefit for switching to 7.  Particularly in the consumer market this will provide incentive for switching from XP or Vista to 7, although it probably won’t be compelling in the enterprise market for at least the first couple of years.  Vista has many good features and 7 appears to be providing all the good parts of Vista without much of the negative.  As such, it should reaffirm Windows’ dominant position in the enterprise.

Implications: Sun has tied its future to open source, but seems clueless on how to monetize open source.  Worse yet, they don’t seem to care!  Meanwhile their core revenue stream continues to dwindle.  Future prospects don’t look good.

Analysis: Sun Microsystems in its heyday had a business model rather similar to the one Google enjoys today.  One major revenue stream, with high margins, funded a multitude of ancillary activities.  In Google’s case the revenue stream is advertising, and it is so profitable Google can afford to indulge in a wide variety of non- or marginally-profitable activities.  In Sun’s case the revenue stream was enterprise servers.  Sun’s enterprise servers, circa 2000, were ubiquitous and carried high margins.  They funded a variety of non- or marginally-profitable activities at Sun.

The advent of multi-core chips is seriously eroding Sun’s enterprise server market share and forcing them to cut margins dramatically to retain the share they still have.  Whereas once it was necessary to spend from $25K to $250K for an enterprise server, today a server with similar power is available as a commodity item for $7K to $20K.  Sun has a successful range of products in this market space, along with retaining their legacy Sparc-based large server products.  However, margins are much lower in the commodity server space, competition is more fierce, and thus the margins are no longer sufficient to fund other activities.  This picture is not likely to get better for Sun.  On the contrary, commodity processors from Intel and AMD continue to grow in power and cores, further encroaching on Sun’s Sparc-based server market.  The recent introductions of the T1 and T2 product line furthers the Sparc architecture, but still competes in the commodity server market.

Sun has storage products, but here they face increasing competition from Network Appliance, Hitachi and EMC.  They are not a leader in the storage market, which means this too becomes a relatively low-margin business as they struggle to survive.

Sun has tied their future to the open source world.  They claim, probably rightly, to be the world’s largest open source vendor.  Swartz is on record repeatedly espousing the view that having the Sun brand widely distributed through open source is critical to Sun’s future.  His philosophy is to seed students, developers and startups with the Sun brand at no charge, with the idea they’ll eventually become paying customers in the enterprise world.  That strategy was used very successfully by IBM in the Sixties and Seventies with their mainframes.  They provided mainframes at reduced prices (or free) to colleges to get students hooked.

The flaw in Swartz’ reasoning, and where the IBM comparison breaks down, is that the products IBM was providing were the same as they sold to enterprises at high margins.  Once the students were hooked on IBM mainframes and moved into enterprises, they paid through the nose to buy more mainframes.  Sun doesn’t have a comparable high-margin product to follow on to their open source offering.  All they have is a subscription service to support those open source products in the enterprise.  That will bring in “some” revenue, but the margins are much lower than in the mainframe and server markets, so Sun as currently constituted will not be able to survive in that model.  Further, there are a lot of choices in the market, and once customers start paying they may no longer use Java or MySQL.  Today’s free MySQL users may become tomorrow’s Oracle users.

Sun’s business model continues to depend on high-margin server business, which is shrinking.  What they’re left with is low-margin commodity servers and open source support, neither of which can support Sun’s current investment in R&D and other activities.  Sun supporters may feel good to know their brand is everywhere, but their future is one of steady decline unless they can either find a high-margin business or re-structure around a low-margin business model.

Implications: Desktop virtualization promises huge savings in TCO, improved security, good user experience, improved manageability, etc.  Despite these advantages, it has many issues that have so far prevented widespread adoption, and likely will always limit adoption. 

Analysis: VMWare, Ctrix and others claim desktop virtualization is the answer to IT's prayers for desktop TCO and security.  Server virtualization has been taking the IT world by storm, so why hasn’t desktop virtualization done equally well?  Is it just because the products aren’t as developed yet?  Is it a public perception issue?  Or are there inherent issues in the model that limit adoption? 

The supporters of desktop virtualization talk about no longer needing to update desktop hardware, better compliance and data security, low TCO for desktops, and the ability to access the desktop image from any machine anywhere.  These sound like compelling advantages, particularly the low TCO and better security.  However, these advantages have been available for 30+ years.  It’s called a mainframe and terminals.  In the Eighties the users migrated en masse away from the mainframe/terminal model, as soon as PCs became available.  Larry Ellison got a lot of press in the late Nineties with his “thin client”, which is essentially the same model.  What has changed now to make the mainframe/terminal (oops, desktop virtualization!) model more attractive?

The short answer is:  Not much!  Users made it very clear in the Eighties that the overriding consideration was the quality of the user experience.  Most users want the independence and control provided by the individual PC.  They want the vast quantities of storage available, the quick response from the screen and keyboard, and the fast (relatively) processing power.  They want these things enough so that all IT’s arguments about TCO and security fall on deaf ears.  Desktop virtualization supporters claim to be able to match the user experience while providing all the benefits of a centralized solution.  Is this claim valid?

There’s not enough space here to fully investigate, but a few claims can be considered.  One big argument for desktop virtualization is to maintain one copy of the OS and push it so all local machines.  This requires a hypervisor running on all local machines that can interface with the OS copy.  That will likely be available at some point, but it’s not generally there today, and few companies are going to replace all their PCs to get it.  Said hypervisor and OS must be able to manage a plethora of hardware, from the very old to the very new.  If that task were easy Windows today would work lot better.  A significant percentage of the problems with Windows are tied to managing a wide variety of hardware.  Desktop virtualization won’t make that problem go away.  It just divides the problem between the hypervisor and OS. 

Backing up local disk storage is a thorn in the side of every IT department.  The total mass of storage is so great it’s virtually impossible to provide enough backup for all of it.  Moving it into a central virtual image doesn’t help – it just moves the storage from cheap desktop disks to expensive data center disks.  That increases cost, which means users have to be carefully limited on the amount they can use.  But that flies in the face of user independence.

Well managed environments today control their endpoints from a central location.  Patch management, updates, virus checking – all these can be centrally managed while still providing users the full experience they desire.  For most users, the incremental gains in TCO and security will not be worth the impact on the user experience.  There is a subset of users, typically those who use the PC for a limited function (e.g., clerks, word processors) for whom desktop virtualization will work and be a fine solution.  For the typical knowledge worker, however, it seems the independent PC/laptop is likely to live on.

Implications: Oracle’s applications have both wide and deep functionality, but that’s not enough anymore.  As younger users move into positions of influence intuitive interfaces and social networking features become increasingly important.  The applications vendor that “gets” this will be in a commanding position.

Analysis: ERP applications in particular have long focused on providing integration and functionality, both things that CIOs and business users have been demanding for many years.  The Holy Grail has been applications that provide every possible function the business users could possibly need, and then tie all those disparate functions together into one cohesive whole.  This is certainly a laudable goal, and remains equally important today, but business users are already starting to move beyond that.

In some functional areas, such as Finance and Manufacturing Operations (not to mention IT) users are accustomed to non-intuitive, complex interfaces.  In many cases the ability to understand and navigate successfully through a succession of complex screens becomes a matter of pride and job security.  In other functional areas, such as Sales, Marketing and HR, users today are less tolerant.  They expect interfaces to be simple, clear and easy to navigate.  This is not an area where the big vendors, such as Oracle and SAP, have been very successful.  Their vast legacy code, combined with the need to provide deep functionality, has made it very difficult for them to provide simple, intuitive interfaces.

Much of the success of the most notable SaaS companies, such as Salesforce.com, has not been because of their service delivery model, but rather because they provided an interface that sales people love.  Business users rarely care whether an application is delivered via a service or a licensed model.  They just want a functional application that’s easy to use.  It’s generally left up to Finance and IT to sort out the delivery model.  Salesforce.com, SuccessFactors and others are making sales because Sales and HR users love working with the applications.  They are stealing sales from Oracle and SAP due to this, and Oracle appears to be finally acknowledging that fact.

At the recent Interop Oracle announced they would be delivering applications with a Web 2.0 focus.  That first of all means a simple, intuitive interface.  Adding still more complexity to today’s overly complex interface will not work, and Oracle realizes that.  That intuitive interface will also provide social networking capability, for example by integrating with Linkedin.  And that’s only the start – once the capability is there viral marketing will likely expand it in ways difficult to predict today.  The Oracle applications will allow mashups, which can be extremely useful in many ways, such as overlaying business data on maps.  Oracle is also promising to deliver a mobile interface for handsets that is actually designed for a small screen, rather than just squashing all the normal information into a smaller font.

It seems questionable whether Oracle can deliver all this within the next 12 months, as they are somewhat promising.  Even if the actual delivery is more like 24 -36 months, however, it promises to give Oracle a remarkably strong position in the enterprise applications market.  Not only will they be a generation ahead of their chief competitors such as SAP, they will also finally have an interface that is competitive with Salesforce.com, SuccessFactors and other SaaS vendors. 

The focus around Oracle for the last 2-3 years has been their acquisitions strategy, so it’s about time the focus shifted back to core applications delivery. 

Implications: Apple cannot win the enterprise market for the Mac, but Microsoft can give it to them.  When the resurgence of the Mac is combined with Microsoft’s continuing insistence on selling Vista to enterprises, the door is open to Apple gaining a solid presence in the enterprise market.

Analysis: Apple’s successes the last few years have almost a fairy-tale aspect.  Few would  have predicted in 2003 that Apple could be where they are today.  Allegedly not even Jobs really believed this kind of success was possible.  Apple’s recent successes have been fueled by their consumer products – the iPods and iPhones, but the Mac has also benefited.  Not many today remember, but in the late Eighties the Mac was considered a viable alternative in the enterprise market, albeit a niche product.  During the Nineties Apple somehow lost that status and by the late  Nineties had virtually no presence in the enterprise.  Today the tide is turning and people are starting to ask for Macs again.  Despite the very legitimate objections CIOs have to bringing in Macs, when a large enough group of users demand Macs every CIO must give them serious consideration.

The pricing of Mac hardware, the fact it’s a single-source vendor, along with the training and support requirements for a new and additional platform are normally enough to eliminate Macs from serious consideration in enterprises.  This picture is changing today, however, with Vista failing to penetrate the enterprise market.  Microsoft continues to adopt a “head in the sand” approach to Vista, with Ballmer recently stating something to the effect that Microsoft would continue to support XP if the users demanded it, but he hadn’t yet seen the demand.  If he hasn’t seen that demand he must have spent the last year in solitary confinement!  Everyone else has seen it.  Vista’s penetration into the enterprise is still abysmally low.

A recent Forrester article said that enterprises should just stop whining and accept Vista, since it was inevitable and they don’t have a choice anyhow.  Wrong!  If this is representative of the thinking at Microsoft they’re due for a very unpleasant surprise in the next couple of years.  Vista’s problems are well-documented:  Excessive hardware demands, limited driver/software compatibility, and continuing reliability problems.  XP is today the enterprise choice, and will continue to be so for the near term.

When given a choice between Vista’s hardware demands, which equate to higher prices for hardware, and the Mac’s typically higher prices there isn’t much difference.  Vista’s compatibility issues with drivers and ISV software effectively offsets the training curve for introducing Macs.  And most convincing – the Mac is a more  reliable platform in the enterprise space than Vista.  Given this list, suddenly the Mac is looking like a viable choice.  Add in popular demand, and CIOs now “can” choose Macs. 

Enterprises will stay on XP if possible and wait for Microsoft to release a new OS that works in their environment.  In that scenario the Mac will gain a little enterprise market share but not a lot.  If Microsoft forces enterprises to go to  Vista, most will comply and simply suffer the consequences.  But a significant minority will revolt and go to the Mac.  Some will also go to Linux – Ubuntu is increasingly a viable alternative.  So Apple can’t win the enterprise market, but Microsoft can hand it to them unless they wake up and provide a viable enterprise OS.

Implications: In terms of volume the iPhone is indeed a “niche product” for Nokia.  However, the impact the iPhone is having on the handset market is much greater than its market share might indicate.  Nokia needs to address their steadily dropping US market share, although the iPhone is only one small piece of that problem.

Analysis: In retrospect, questioning the Nokia CEO about a slipped product launch at the same time he announced disappointing earnings and forecast was unlikely to elicit a positive response!  It is not surprising that under the circumstances he would brush off the success of Apple’s iPhone as insignificant.  It is indeed true that in terms of worldwide market penetration the iPhone is a niche product, so his statement is factually accurate.  It is also important to keep in mind that while most of the media in the US has an understandable US bias, the Nokia CEO considers the US as only one among many markets.  An important market, to be sure, but arguably no more important (possibly even less) than Europe or China. Finally, while the iPhone has been very successful in the US, it has not generally matched that success outside of the US.

Some people look at the iPhone and compare it to the iPod, which has come to dominate the music-player market.  The handset market was much more competitive when the iPhone was introduced than was the music-player market when the iPod was introduced.  It was never feasible that the iPhone could come to dominate the handset market.  A better comparison is probably the Macintosh computer.  The Macintosh never dominated the PC market.  At best it was never more than a “niche product” in terms of market share.  Yet the impact the Macintosh had on the PC market was huge.  Much of the impetus behind today’s Windows was the Macintosh.  It’s impossible to say what Windows would look like today (or if there would even be Windows!) without the Macintosh, but certainly it would be very different.  The iPhone seems positioned to have a similar effect on the handset market – although it may never dominate in terms of volume, it may well dominate as a design driver.

Nokia needs to address their slumping market share in the US, which by some accounts has dropped from 20% to 7-8% in the last couple of years.  Releasing an iPhone look-alike will help that, but the key issue seems to be their rocky relationship with US carriers.  Since carriers effectively control the handset market in the US, Nokia must improve that relationship to regain market share.  Along the way they need to get their iPhone look-alike out the door.  Early reports suggest that it still needs a lot of work on the user interface, which is the key to the whole product.  The fact Nokia has now slipped its release to later in 2008 suggests they understand the problem and are working to resolve it.

It is very unlikely that in five years the iPhone, or its successors, will dominate the handset market.  Nokia’s market domination worldwide (40% market share) is likely to continue.  It seems very likely, however, that in five years the influence of today’s iPhone will be strongly felt throughout the entire handset market.  Not all the products are going to look like today’s iPhone, but the best features of today’s iPhone will be integrated in with the best features of other handsets to create an overall superior product.  Apple’s products don’t dominate most of the markets they sell in, but their design influence does.  The iPhone’s design and features will forever change the way handsets are viewed.