The Battle of The Giants

It had to happen. One of the outcomes of most recessions is that there are some fearfully large consolidations of big companies - and yesterday saw another big announcement in a market sector that is heating up big time.

Hewlett Packard announced it was acquiring networking company, 3 Com, for around $2.7bn in the same breath as producing better than expected profits on lower sales last year. For those in the industry, this the annual musical chairs time at the computer giant, HP, and many will find that a few more chairs than usual have been whipped away as they keep making sure they are as lean as possible.

Ostensibly, the acquisition is to give HP a footprint in the lucrative Chinese market where 55% of 3 Com's sales come from. However, there are other motives as there have been major moves by networking giant Cisco to move into the server market at the datacentre which is a direct threat to the giants of that market, HP, Dell and Sun. The 3 Com acquisition, in that context, is all about augmenting HP's portfolio of networking switches and infrastructure devices to ensure they have all the pieces of the server, storage and networking products required to maintain their market position as Cisco put their formidable weight into this market.

There is more. Cisco's market play into the server market takes it from its luxurious, high profit network business where it dominates into the grimy and lower margin world of PCs and servers where HP have great skill in thriving. It could be argued that Cisco will find it hard to apply its typical cost models in the server market to be able to compete or make a decent profit. HP's move is certainly a clever defence of their position, although in reality, 3 Com has only a single digit market share in Cisco's core markets and was no longer the big player it used to be. Some analysts feel that companies like Brocade would have been a better buy, but in the great scheme of things, any old company in networking would have done.

HP have learned how to acquire and absorb big companies very quickly. Starting with the huge takeover of Compaq which cost Carly Fiorini her job, they have more recently acquired EDS which almost doubled their workforce. Meanwhile, their arch rivals in the PC and server market, Dell, has been building out its portfolio to compete by adding Equalogic and Perot into its midst making it a much stronger company.

What does this all mean for companies, large and small? Well this period of formidable consolidation will likely increase the massive competition in what is already a highly competitive market for PCs, servers, storage and networking. It is likely to drive prices down further which may be a good thing - but for those companies who sell these products, the decreasing profit margins in selling solutions based around these companies and Microsoft is squeezing the life out of what was a vibrant market to be in. I don't think this will make life much better for them.

It certainly means that Cisco and HP are now toe to toe and head to head in the market. It will be a battle royal to win the loyalty of the customers.

One Way To Alienate Customers

I mentioned some while ago that the Sun-Oracle proposed hook up was going to be an uneasy marriage. Their latest marketing effort illustrates how silly this can get.

On the back page of The Economist is an expensive advertisement that proclaims that the performance of Oracle working on Sun machines is better than the same software running on IBM machines. Thus, they proclaim that ‘Sun + Oracle is Faster’. The proof of this claim rather bizarrely will be published or at least be available from 14 October. It does not actually say which October this will be, but we can only assume that it is this year. IBM, meanwhile has about one month to respond to this claim and before they twist the argument, this is as compared to an IBM Power 595 Server Model 9119-FHA to be exact (plus a few caveat figures to prove they actually ran a test) but they do point out in the same small print that this model was available back in December of 2008. So Sun and Oracle have had around 9-10 months in order to beat the performance of that IBM machine and have not considered any subsequent models in between. Oh, and the model, or at least the solution package, is presumably not available until 14 October of some indeterminate year in the future.


So what does this advert say to Oracle users? Not much really. The software is not getting any better and that you need a damn powerful machine to get performance out of it. And if you bought an IBM server to run it, then you are a chump, basically. Yes, in the world of negative advertising this takes a good bite of biscuit. You can infer from the new line of advertising that there must be some kind of ‘tuning’ going on by Oracle to make the behemoth software (they are the second largest vendor of software worldwide with over 50% of their market) run slightly faster on their now home brand hardware – making any other type of Oracle user at a disadvantage. The advert does not mention Dell or HP or any other brand of hardware but one can infer that if the software is now specially tuned for Sun then users of those other brands will also be disadvantaged for the future, if not by 14 October.

What would happen if Microsoft adopted the same approach? It would alienate just about every other manufacturer of PCs should they throw their lot in with, say, HP. Immediately, all those years of careful building of an independent software brand that runs on any x86 chip at speeds controlled by the manufacturer (but the user gets a consistent look, feel and features no matter which one they choose) would be lost as Microsoft would be saying, ‘run it on any other brand than HP and you are getting less than your money’s worth’.

That’s effectively what Oracle has done. Many market analysts saw this as a natural move by Oracle to fight IBM, but it just erodes the software vendor’s independence on hardware. Whether they like it or not, there is a vast world of business users out there that have chosen platforms other than Sun to run their software – now they would really have to start questioning whether their software company is committed to their end user customers or ONLY those end users who run Sun hardware. And remember, Sun were struggling prior to this $7.4bn proposed takeover.

From a customer perception, this advert is the first major manifestation that Oracle has ditched its independence and will be preferentially developing software tuned for the Sun platform only in the future and that there is even now a risk that subsequent versions of Oracle will only run on Sun. So if you are an Oracle user today and not using the Sun platform, perhaps this advert poses two questions rather than the one intended: 1) Should I buy Sun in the future and if not, 2) Should I be reconsidering my choice of software for the future?

In the world of large scale software sales, while you can have umpteen nice OEM deals, latest offers and special relationships with lots of hardware vendors but, ultimately, the last thing you want to do is to lose a deal because of hardware allegiance. Better to make sure they buy your software because it best fits their needs rather than the needs of the hardware they run.

Maybe I am being pedantic here, but I found this advert astounding – then again, I thought the Sun-Oracle tie up a big mistake and smacked more of desperation by one or other of the companies rather than of real strategic thinking.

Sunny, With Dull Patches

So Oracle has bid $9.50 a share for Sun Microsystems, valuing the US hardware fallen giant at $7.4bn. It's a far cry from what Sun might have been worth a few years ago but it currently is worth a 42% premium over Friday's closing stock price for Sun - how the mighty have fallen.

IBM had tried to secure a deal with Sun last month but nothing came of it. To be frank, it may have been a more natural fit in some ways - two hardware vendors with software and services interests combining to plug the gaps in each others' portfolios. But the recent past has seen the once mighty Compaq get severe indigestion when it took over the lumbering has-been Digital only to see the failed combination being bought by HP under Carly Fiorini.

The fact is, Oracle is a software giant and although threatened by the likes of IBM, in reality the two models are poles apart. The computer industry has always segmented the business models of software and hardware. Many hardware companies had software interests but they never really came to the fore and dominated any specific space. Similarly, some software companies have had some hardware interests but again, none that ever really dominated a sector.

The plain facts are that software is a gross margin rich business model with development costs underpinning it. Hardware has seen its gross margins eroded, made worse as each vendor has attempted to service major supply agreements as Prime Contractors and therefore had to buy other hardware vendor products in at reseller type margins, often securing long term supply at single digit margins on other vendor equipment and so dissipating its true gross margin on deals. It has been the nature of the beast in getting revenue growth at the cost of profit. Dell, in particular, has felt the pinch using this tactic.

So what would happen when you combine the two models, particularly of two giants of their specific sectors? I have to say, that the jury is out for me. Sun made a walloping $1.9bn loss last year on £13bn of sales but Scott McNealy, the colourful Chairman behind Sun, reckons that Sun will contribute $1.5bn of profits this year and $2bn next year to the new group.

Industry Speak

Industry analysts reckon that Sun's customer list will give Oracle access to those customers who are not currently using their database products. How often have we heard that kind of speak before? The fact is that hardware vendors rarely 'own' their customers in the way that a software company does unless they have a major stake in their applications or infrastructure. The chances of Sun dictating what database it customers use are pretty slim to say the least. Moreover, it is not as if Oracle is not known in the corporate world so that Sun customers will not have heard of them - all in all it's a silly argument and would be similar to Microsoft buying Apple to force their operating system and products down the throats of Apple customers.

The best way to make such a combination work is to produce database-engine boxes which were tuned in performance to run Oracle applications - 'Appliances' if you like. The storage industry has a number of such products and giants in this sector like EMC and NetApp have acquired strongly tuned software products to work specifically with their hardware.

But such an avenue is fraught with danger, as EMC has seen in acquiring VMware. As great an acquisition as that was (EMC got its money back by floating just 10% of VMware stock), the issue was always that VMware had to remain independent of its parent in order to remain the credible market leader, particularly with Microsoft gearing up to get on the attack. If VMware were to favour its parent by tuning performance or showing preference, long term it could ruin its position with the likes of HP, IBM, Dell and others and so allow Microsoft in as the 'independent vendor' who was hardware agnostic.

This is the danger for Sun-Oracle. If Oracle tries to leverage its new hardware purchase by tuning the performance to the hardware or showing preference in development, then the vast array of hardware vendors will slowly but surely get jaundiced and it allows a gap for an independent supplier with good brains and products to win their hearts as there is no vested interest. Right now, HP or IBM executives who have a strong relationship with Oracle, are probably thinking that all their Oracle installations are potential Sun Microsystems targets.

Long term that may not be proven to be the wisest of choices. If you are going to buy a hardware vendor as a leading software supplier, make sure you buy the biggest as your first step, not the cheapest, as you may just have bought yourself not just a chump but a liability that will lose you a lot of friends.

What may have started as a plan to strengthen Oracle's position against IBM may be the first gong of the death knell of a giant of the software industry. If there is one watchword in this industry that determines success it is FOCUS.