At VMworld 2011, VMware CTO Steve Herrod gave a compelling roadmap of how the company is forging ahead.  If you look under the covers, a lot of the underlying technology and consequently growth, for VMware comes from a series of acquisitions over the last two years.  Take a look at two exhibits below:

Here are some questions that are worth considering:

1. Acquisition pattern and strategic intent: Is there a pattern in how VMware went about acquiring the companies in the picture? Did managers at VMware formulate a strategic intent on how they were going to offer a vision of VMView that is now (or at VMworld 2011) being offered, back in 2009? Or did they just stumble into it?  In other words,  was this acquisition picture carefully orchestrated, or simply a series of experiments that created a coherent picture much later?
2. Acquisitions, growth and innovation:  For technology companies, are acquisitions the only way to grow?  How can a technology company that serves enterprise users innovate continuously?  Is that innovation coming out of acquisitions only? (what about scaling?, integration?)
3. Leadership change: Did a CEO change contribute to any of the change in VMware’s growth? 
4. What can managers facing disruptive forces do in light of 1, 2 and 3 above?

I don’t have answers to these questions (because I’m not privy to how managers at VMware made these decisions!), but I’ll offer my point of view as an observer.  In this post, I’ll expand on question 1) and follow up with another for the other two questions.

Zig-zag?

First on the acquisition pattern and strategic intent:  If you had followed VMware pre-Aug 2009, they had only acquired 6 companies until that time, and all of them were directly related to their virtualization technology, except possibly one (TRANGO Processors) which was about running VMs on mobile phones and even that was too close to virtualization.

Until VMware acquired SpringSource, there were no significant events in core enterprise technologies (read: impacting developers) for VMware.  Most strategists at the time VMware acquired SpringSource would have thought the natural adjacency for the hypervisor would have been a management technology for the hypervisor, not a developer framework.  But VMware declares its intentions for a bigger vision by acquiring a development platform.  Or did managers at VMware really know Spring was going to serve VMware’s PaaS forays today?  In fact,  even managers within VMware would have questioned on why invest in something so far from what VMware was already doing is good.  That would have been a tough pill to swallow - managers would have after all wanted investments driving their current product lines not a new category.  You’ll see over this period of active acquisitive growth, VMware goes about collecting pieces on the chess board small, unrelated, yet opportunistic assets as some might say.

What does VMware do after acquiring SpringSource? You’d think immediately after acquiring SpringSource, VMware would go and acquire a bunch of technologies close to SpringSource to “complete the stack” for developers.  Actually, VMware didn’t do that at all.  It switched all the way up and out to a hosted email service (Zimbra) next.  That acquisition was puzzling to many watchers.  That made no sense to the casual observer.  

SaaS? Developer? Operations? All of the above?

Did VMware want to play in the SaaS marketplace?  Did VMware want to get into the collaboration software business?  If that were true, you’d think at this point in time,  they’ll go and buy a set of SaaS assets to play in the desktop application space.  Not until two years after (i.e. just a few months back), did they acquire three more SaaS assets (SlideRocket, DigitalFuel and SocialCast) to make a “starter pack” for enterprise SaaS. 

 Is it possible that Zimbra was an “experiment” for VMware?  To test the application space? Maybe Yahoo’s divestiture was just a serendipitous moment for VMware to enter the application space so far away from the hypervisor.  Yet, today, it looks so naturally close to it.

After acquiring Zimbra, VMware acquired a whole bunch of management technology assets from parent company EMC.  My guess is, around that time virtualization was already getting to the tipping point of large scale adoption, and management of virtual machines was becoming a problem.  So VMware changed gears, and acquired management technologies to beef up its vCenter portfolio.

Soon after VMware moved further down the stack (from management) and acquired Rabbit MQ.  Why?  AMQP was emerging as the best protocol for messaging in middleware and Rabbit MQ implementation was the perfect integration that someone who already had Spring would look for to broaden enterprise adoption of Spring. 

But wait.  If you thought VMware was paying attention to developers only then  you are wrong.  Suddenly cloud became the buzz word everywhere and security the #1 concern to cloud adoption.  What does VMware do?  Strengthen security with TriCipher and NeoAccel acquisitions.

Again, VMware switches back to developer needs acquiring an IDE (WaveMaker).  

Enterprise Collaboration SaaS Starter Pack (yes, that’s a lot of words)

Finally VMware goes back to actually completing a simple usable starter pack, it started almost 2 years before with that Zimbra experiment. So SaaS starter pack = Zimbra + SlideRocket + SocialCast – which can, conveniently be bundled with their resurrected VMView at VMworld 2011.  Pretty impressive, eh?  At VMworld, I chatted with Javier Soltero, CTO for SaaS, App Services at VMware and he asked me (I should say, gleefully) “does it all make sense now?”  Of course, it does!  In fact, my colleague Cote has a “we get it post” aptly titled The VMworld 2011 Application Story.

Deliberation vs Experimentation

But did managers at VMware set out on this journey with deliberate intent? Was there strategic intent at VMware in August 2009 to be where it is today? I would argue it doesn’t look like that at all.  The portfolio looks like a zig zag map that got filled out progressively (and will continue to) and the puzzle appears to be getting solved.  Well, sort of.

Last week, I had the opportunity to speak with Michael Dell – himself an incredible experimenter – no less than creating a $61B global technology leader that he started out of his college dorm room.  I asked him what drives good strategy.  His response was direct – two words – “listening, learning”. 

If you look at VMware’s acquisition strategy, it seems to fit that description.  But do we know that?  Is VMware adopting an Agile acquisitive strategy? (small experiments, little bets, listening and learning, iterative continuous improvement) - More on all of this in post#2.

The No-Server Summit was initiated to understand the dynamics companies that are starting-up, thriving and ultimately staying in the cloud.  The process included asking questions like:

• How do they manage their operations and is it a permanent form of operations?
• Are they in transition to better known traditional computing models that exist today?
• Is this a new paradigm with applicability beyond young and small companies that can be better understood?

What sets the “No-Server” organization apart from traditional organizations, is that the company will have a firewall and wireless access points, but no servers or other traditional datacenter hardware or infrastructure.

In reality, there isn’t one community but at least three groups creating companies (individual departments and product groups also, let’s collectively call them “Entities”) that have adopted this type of infrastructure.  For those of you thinking of only bleary-eyed developers looking for a test environment for their code, you may be surprised at the breadth of other use-cases and users.

• NewCos and Venture Portfolio Companies
• Products-with-significant-Software-Feature Entities
• Web and WebScale Entities

In general, each of these Entities is seeking some mix of the following benefits of Cloud:

1. Cost Savings - Companies starting in the Cloud, Lean Startups seeking OpEx and CapEx savings
2. Product Agility – Software Application Entities requiring product agility (DevOps)
3. Infrastructure Agility, Simplified Operations and Management (also #1 & #2 above)- Entities with demand changed driving scale or elasticity manipulation
4. Acquire Advanced Security/Performance/Uptime (also #1 & #2 above)-

The Entities that are adopting Cloud each have their own combination of qualities that are vital for its own success.  NewCos and Venture Portfolio Companies find all four Cloud Qualities as valuable, some to more of an extent than others depending on their business model.  Products-with-Significant-Software Feature companies tend to be almost solely interested in Product Agility to create customer delight and beat their competition.  While Web and WebScale Entities tend to favor the Simplified Operations, Infrastructure Agility and Management qualities of the Cloud.

Cloud Qualities – Cost Savings

The term “Lean Startup” is a term coined by Eric Ries, who uses the term to talk about simple and efficient creation of new ideas, products and entities that manage them.  Eric’s original work was focused more on lean product innovation rather than infrastructure, none-the-less the extension to lean entity creation and infrastructure operations was planted and has grown.

Lean company creation is also a generalized concept now mandated by nearly all of the Venture Capital community who are directing their start-ups to the Cloud for IT.  The VCs see a very large advantage in limiting the CAPEX of a firm in soon to be outdated servers, networks and storage infrastructure.  And although the cost side of the equation is appealing, the firms also gain near-instant provisioning of IT infrastructure and minimum IT maintenance effort which can lead to faster time-to-market.  The firm 8K Miles has made this very set of companies the sole focus of their business.

Cloud Qualities – Product Agility

Product Agility is another principal reason for Cloud operations.  And while Cloud infrastructure allows for nearly instant IT infrastructure, it is not that attribute, but the iterative operational-construct of DevOps that is the most valuable component of Cloud infrastructure adoption.

What is DevOps?  Simply put it is a new way of writing application code and running a company’s infrastructure in one construct.  This construct, as the name implies, the combination of the formerly silo’ed operations of the Development and Operations groups.  What DevOps is, will be discussed in later posts.

National Instruments is a manufacturer of physical Field Programmable Gate Arrays (FPGAs) that are used in a wide variety of use cases including the measurement of physical environment qualities such as humidity and temperature.  The company not only has a physical product with no relation to the cloud, but it also maintains its own datacenter running ITIL operations.  However, one day, an individual product team decided to create its new product features using cloud development and located those services in the cloud as well.

The decision made by the product team to go to the Cloud initially included each of the four Cloud benefits (above) including the fact that the environment was simple and easy to acquire.  However if the team is asked today, the overwhelming highest value comes from the agility created through the process of combining Development and Operations (DevOps).  It is this single quality that has allowed the firm to create features at a much faster pace than other departments, divisions and competitors.  So much so, that other divisions are also now asking for the same capability.

Cloud Qualities – Infrastructure Agility, Simplified Operations and Management

Infrastructure agility is also a key quality for many firms when adopting a No-Server infrastructure.  Elasticity, scale and agility of the platform supporting the entities application becomes integral in allowing the operation to handle the changes in its business and change in volume.

Animoto is one of the poster children for Cloud.  The company launched its initial product in 2007 and then later Animoto launched their own Facebook application.  The Facebook application went viral and it became a popular way for users to showcase their photos on the social network. Within a week, the company had to increase the number of servers with Amazon’s Elastic Computer Cloud (EC2) from 50 to 5,000.

Soon afterward, Animoto became one of Amazon Web Services’s cloud computing success stories. By mid-April, Animoto’s Facebook application had nearly 750,000 people sign up in three days. At the peak, almost 25,000 people tried Animoto in a single hour.  Amazon Web Services’s ability to meet the demand of Animoto’s viral Facebook application resulted in Amazon.com’s eventual investment in Animoto.  

Netflix recently closed the last of its physical datacenter operations and now has all of its operations in the Cloud.  This decision was made not prior to a new volume in customer visits, but in preparation for the increase.  Specifically, Netflix is using the flexibility of Cloud infrastructure to meet its own expansion strategy to expand its streaming product around the globe.   The plans being so aggressive, that Netflix could not build facilities at the pace of the planned expansion.  In addition, the agility of the DevOps method of operations also is a major factor allowing Netflix to push new code and product features many multiples of times each and every day.

Cloud Qualities – Acquire Advanced Security/Performance/Uptime

Although the Cloud is still subject to mortal attributes, for many of the No-Server firms it is an individual Cloud hosting company’s scale of operations that allows it to be vastly more secure and available than any individual team of IT generalists.  Much more so than any individual development team or department can muster.

Summary

In the end, cloud operations are sought for several reasons: Product-Agility, Scale, Elasticity, Simplicity, and Cost.  To any one firm, a single one of these features may be the driver for adoption while the other attributes beyond their favorite do not matter at all.  As a result, each feature can drive adoption on its own.  The others are just gravy.

For most of these firms it seems, once their IT operations are in the Cloud there does not appear to be a point at which they will return to physical datacenter operations.  In fact, if DevOps was adopted for development in the process, they almost cannot return to traditional IT operations and infrastructure. 

Having stated the one-directional trend of Cloud adoption of No-Server entities, there are still a select number of companies that have pulled their operations from the Cloud.  We will discuss those and their reasons for the exit in a later discussion.

A few months back, the three of us organized over half a day of discussions at Tech Ranch Austin around a simple idea - what happens when you operate your business with ‘no servers’.  Like Marc Benioff who talks of ‘no software’, we explored ‘no servers’.  We invited 25 CTOs and CEOs from bleeding edge startups in Austin and a few from out of town.  They were kind with their time and happy to share their experiences and insights.  Over the course of next few weeks we plan to post what we learned from the No Server Summit.

Presented below are our initial hypotheses and reactions:

Cloud is real, rain or shine, it is here to stay:

One thing we were curious about is, is there an inflection point when owning servers in-house becomes advantageous over renting servers – from cost, manageability, security or any other reason.  Was the move to the cloud simply a question of easier, low risk cap-ex to op-ex model for the short term or is there much more to it? The resounding answer we got was, there was no turning back from the cloud for these startups.  One start up just signed a multi-million dollar deal with a financial institution that is demanding Service Level Agreements much different from the startup’s previous customers.  The startup is working on meeting those SLAs (both uptime and security) but within the cloud, not outside of it (i.e. not by adding an in house datacenter).  Another is offering a HIPAA compliant product for outsourcing pharmacy servers to the cloud.  That means pharmacies will not be buying servers as more of these products enter the market – an interesting future scenario we need to think about.

Agile development is driving cloud adoption which in turn drives future opportunities for growth: 

Wayne Gretzky famously said ‘skate where the puck is going’.  In cloud computing terms, the puck is going to PaaS.  Developers aren’t developing on the old SDLC waterfall model in Java or .NET.  They are using web scale applications and platforms right in the cloud – because that is what supports their iterative, fail first, fast, lean approach to software development.  The CTO of Dachis Group Erik Huddleston, gave a fantastic presentation on Lean Startup meets Lean Development. 

DevOps will become a major user group in IT organizations because of cloud computing specific needs:

What we gleaned from these dozen plus startups is, none of them had a developer or operations silo like traditional IT organizations.  Because of agile, and because of the necessity to understand application infrastructure in real time without the underlying hardware infrastructure workings, developers are able to take up more operational roles, and traditional ops is getting less relevant in these future organizations.  Imagine what kind of product offerings we need to come up with to target DevOps.  What this also tells us is, the traditional systems management vendors are going to have a tough time cloud-washing or selling their wares as they are.  Damon Edwards from DTO Solutions kicked off the discussions around this topic with a great presentation.  URL: http://www.slideshare.net/dev2ops/devops-it-operations-as-a-strategic-weapon 

The SaaSification of things

You’ve probably heard of the “internet of things” – where everything around us be it the water faucet or the milk carton in the fridge, is connected, monitored, measured and controlled with a chip in it.  If the internet of things is the shape of what we experience, the SaaSification of things is how we get to that experience.  Case in point  – National Instruments, a 30 year old 5,000 employee company in Austin, involved in design, prototype, and deployment systems for test, control, and embedded design applications is moving its FPGA compute offerings to the cloud.  Even vendors with solutions that don’t directly have a SaaS model are moving parts of their solutions to the cloud with hybrid approach.  Speaking to the audience, Ernest Mueller from NI says the cloud not only reduced cost, but actually improved up time and security compared to NI’s own datacenter.   Full presentations here: http://www.slideshare.net/mxyzplk/business-model-driven-cloud-adoption-what-ni-is-doing-in-the-cloud
http://www.slideshare.net/mxyzplk/why-the-cloud-is-more-secure

If there’s anything we learned from this 5 plus hours of discussions with entrepreneurs, it is this – the nature and rate of change we expect to see in the next few years is unprecedented.  We can brace for it or be left behind.  

Attendees:

Scott Johnson, President, CopperEgg

Toan Duong, CTO, 80legs

Alberto Alonso, CEO, Global Gate Systems

John Hill, Sr. Group Manager, LabVIEW Cloud and Web Services, National Instruments

Ernest Mueller, Cloud architect, National Instruments

Ned Hill, Managing Director, DFJ Mercury, Houston

Rob Roberts, President, Pervasive Tags

Damon Edwards, DTO Solutions, San Francisco

Dan Greff, CEO, Cloudmasons

Daniel Kim, CEO, Mindquilt

Israel Gat, Director, Agile Product & Project Management Practice, Cutter Consortium

Shion Deysarkar, CEO, 80legs

Tom Bishop, CTO, StoredIQ

Eric Anderson, Co-founder, CopperEgg

David Perdue, VP Engineering, CopperEgg

Erik Huddleston, CTO, Dachis Group

Cody Powell, Co-founder, Famigo

Dave Michaels, Director of Business Development, Tech Ranch

Andy Land, VP Marketing, UnboundID

Jim Stikeleather, Chief Innovation Officer, Dell

Chad Walters, VP Engineering, Socialware

Kevin Koym, Managing Partner, Tech Ranch Austin

Matt Wolken, Corporate Strategist, Dell 

Michael Coté, Analyst, RedMonk

Prabhakar Gopalan, Sr. Corporate Strategy Consultant, Dell

Finally pulling together application-layer strategy. What “Post-PC” really means.

Ever since the infrastructure super-star bought up-stack into applications with Zimbra, people have been head-scratching about VMware’s long-term strategy. Here was a company killing it with the Morlocks trying to get a piece of the Eloies above: what did the kernel-wizards know about UIs and end-users? There’ve been other buys like SlideRocket that have been the butt of jokes in the Morlock crowd. About a year ago, I was briefed on something called Project Horizon, and while the goal of putting together SSO and a private app store (to simplify it) seemed wonderful, the prospects for that kind of good idea in the staid world of corporate IT tend to be poor.

Well, kudos to VMware for not being as cynical as myself. This year they pulled it all together in a great day two keynote focused primarily on enabling new ways to manage knowledge worker’s application needs. Done up with mock post-it notes and white-boards, VMware laid out a credible vision for the future of applications. There’s a long, risky slog to make that vision work, but they’ve thrown together plenty of pieces that’d be required to make it eventually work.

The Nature of Applications Changes

In summary, the big bet is that most applications will be delivered as more of a SaaS or at least hosted application than a desktop-bound application. While we’re passing the time on the way to the future, you’ll still have plenty of desktop-bound, applications, but like a web-faced green-screen application, those traditional applications get isolated and wrapped up in VDI. More than likely, as the cliche goes, those applications never completely go away, but they lesson in importance and, thus, budget and attention.

Thanks in large part to Apple’s mind-share in IT (of both types: consumer and corporate), VMware finally has the problem for the solution they’ve been building. Namely, people don’t want to use traditional IT to do their day jobs. As I’ve obliquely argued recently, they want to use their own, hand-selected IT stacks - users don’t really think of it as “IT,” but it’s useful to label that collection of employee owned and selected services and devices as such.

Consumer IT Experience Eclipses Corporate IT Control

More importantly, and threateningly to traditional IT, consumer IT tends to work better than corporate IT. Worse, there’s nothing but corporate policy to stop employees from using consumer IT - yeah, and you can guess how much employee care about the finer points of policy after years of clicking through 20+ page iPhone user agreement screens. Corporate IT becomes a nagging nuisance to employees in these situations.

One of my current theories is that consumer IT is so much better because consumer tech vendors are forced to focus on the end-user. Keeping the end-user happy and engaged with the technology is profitable. The end-user experience a huge differentiator in such businesses. In contrast, corporate IT vendors long dropped end-users as their primary concern and instead focused on control as the primary feature. Focusing on control effectivly means disempowering the end-user, if only through crappy UIs and ancient feature-sets.

Getting back to VMware, in this context, a traditional IT vendor (like VMware) has a choice to make: do I fight against consumer IT and appeal to my base to maintain current revenue and customers…or do I embrace the end-user and hope they’re the future path to profits? As company like VMware, given corporate will, can do both. Cash isn’t the issue, it’s building up the corporate will to build tomorrow’s disruptive model for applications.

Whether it’ll work is up in the air. It’s very early, and I don’t think we can even sort out some of the basic theories that’d allow a prediction. For example, will “open” platforms win out (as with the first generation web), or will vendors be able to control the ecosystem (as with Apple and Facebook)? That said, I’d put my money on embracing consumer IT rather than traditional IT models.An evermore computer literate workforce isn’t going to put up with the usability and functionally-anemic horrors of corporate IT. And, really, it’s not clear that corporate IT should be the one spending resources to provide such services when others are seemingly doing it just as well, if not better. As example, you can point towards Google Apps for email and calendaring, dropbox for file-sharing and data backup, Evernote for information management, and services like Yammer and Chatter for collaboration. VMware even announced a project to make a “safe” dropbox, Project Octopus…but, again, when we already have a bird in hand, how much is another bird nesting in a stack of policy documents in the bush worth?

The Meaning of “Post-PC”

Looking at the bigger picture, there’s a second-order effect going on here, a stealth strategy if you will. The term “post-PC” has always bugged me. Smart phone, desktop, tablet, or laptop - they’re all just “computers,” really. That’s why I liked Microsoft CEO Steve Ballmer &co.’s recent characterization of tablets as just another PC (and the resulting analysis from Horace is one of my current favorites from asymco-land). But, I’ve realized now that when people say “post-PC,” whether they know it or not (most do, I suspect), they’re actually saying “post-Microsoft.” Think about it next time you come across the phrase “post-PC”: where does Microsoft fit in to that vision? And while users of that phrase, like VMware, make plenty of here-and-now money with Microsoft, the futures they usually aspire to have little to do with Redmond.

Clearly, Microsoft sees this and I’d suggest that it explains their aggressivness in cloud, web, and even opening up their platforms to non-Microsoft technologies like Java and ruby. Microsoft’s strategy now, more than ever, has to be retaining current customers by migrating them to new Microsoft-favored clouds, like Office 365. At the same time they’ve got to not only come up with something new in the way applications are provided and used, but commerzile it broadly (as a recent example of such a thing: mobile apps to Apple and Android). It’s not good enough to just “migrate” to a new form-factor while the likes of Apple, Googe, Salesforce, and now VMware are trying to do much more: namely, change what it means to be an “application” and how applications are used day-to-day.

That’s the chalenge the industry as a whole has now: (re)focusing on users and giving them innovative, new ways of solving their day-to-day problems productively. My sense is that rather than the current ideas of stability and control, everything IT does (and there’s a lot still) revolved around that: empowering the end-user to do their job better.

(Originally posted at CoteIndustries.com)

(Source: )

Two of us are off the VMworld next week in lovely Las Vegas. We will send dispatches from the field of the announcements, whatnots, and interesting people we come across.