5 (more) reasons VMWare need to be careful

I've written in the past about VMWare and how I think their bubble may well one day burst if they aren't careful. The overriding reason has been the general level of arrogance whilst making their, admittedly impressive, way up to the upper echelons of software company ranks. It's no secret they wouldn't quite be top of Steve Ballmer's Christmas card list, and their duel with Citrix also seems to hot up quarter by quarter; indeed they also had a go at Google at the recent VMWorld expo, claiming the cloud in cloud computing should be a private one rather than the public internet.

I recently received a very insightful piece of work from a San Francisco based financial analyst I speak to quite regularly, outlining three very distinct reasons he thinks VMWare need to watch their back, quite apart from the fact that Microsoft so obviously have them in their crosshair (and how many companies in history have survived that for long?) I obviously can't take any credit, my contact has kindly allowed me to reproduce it.

Reason no. 1: ELA over-selling

"Server virtualization continues to have great interest from enterprise customers as they save money with server consolidation and gain benefits in areas such as a more flexible and resilient environment. However, we pointed to VMWare's license billings (best top-line metric) since before the company went public. We had noted as early as Nov 2007 that ELAs were a significant influencer of growth (and then deceleration) for VMWare's numbers."

So VMWare have been overselling to large enterprises, bolstering figures as best they could, for as long as they could, using Enterprise Licensing Agreements. These are now naturally starting to dwindle and are probably having a large impact on VMWare's retractions.

Reason no. 2: Core density

"As core density per processor increases, they are able to support more workloads. VMWare charges per processor pair (which may have multiple cores) and not per core, although there is a limit at the current time at 6 cores per processor. VMware will likely again benefit when core density makes the jump from quad core to octocore, which will occur shortly, but dual and quad core will likely remain the standard for some time."

Are advances in hardware and chipsets having an impact, bearing in mind how VMWare licensing works?

Reason no. 3: Asset-sweating

"As customers are becoming more experienced with the technology, they are able to drive consolidation ratios higher. This is due to buying servers that are optimized for virtualization as well as optimizing network and other infrastructure configuration for virtualization. Also, customers become more comfortable with the characteristics of their workloads and better understand the
server utilization profile for a given workload."

No surprise in this sort of economy, customers are pushing their VMWare implementations as much as they can. As any IT vendor Powerpoint presentation of the last 15 years will tell you, everyone will help you do more with less - so is VMWare becoming its own worst enemy? Perhaps their technology is becoming too good for their own good!

I'd like to add 2 more reasons myself. The first reason is, of course, the troubled time in which we find ourselves, however this is affecting everyone; it is neither VMWare's fault nor is it anything they can really do much about. Their technology is in fact such that it will fare much better during a recession than other vendors' but this won't protect them completely from the huge cutbacks in spending. In a technology survey published by Goldman Sachs last week, server virtualisation slipped from number one to number three in relevance this year. The new number one? Cost cutting.

And finally there's Citrix. The article below announcing XenServer was to become free of charge got me in hot water, I was too quick off the mark, but my enthusiasm was surely understandable. I have heard descriptions of this move being "game-changing" and I couldn't agree more. How can a large organisation possibly now just go out and buy VMWare without trialling Citrix and Microsoft too? And then they must surely provide very good justification as to why those two solutions aren't up to the task (which, in Citrix's case, it most certainly is - as the latest case study at Tesco proves). How can VMWare continue to justify their extortionate prices when their two main competitors are giving away a large percentage of their capabilities? If Citrix don't go on to profit off the back of this momentous decision, they will have certainly achieved one thing - ensured every VMWare sales guy on this planet is going to have to answer some very uncomfortable questions.