Thursday, March 19, 2009

Sun's Cloud (4 of 4)

This entry has been cross-posted from Jonathan Schwartz's Sun blog.

Sun's Cloud (4 of 4) - In the last three updates to this blog, I've tried to set out a clear direction of where Sun's headed. I've talked about our three basic priorities:
  1. Technology Adoption
  2. Commercial Innovation
  3. Efficiently Connecting Adoption and Commercial Opportunity.
I'm hoping you've got a clear picture surrounding the first of these two priorities - how and where we drive software adoption, and focus our commercial efforts.

So now I'd like to talk about the linkages - while also addressing one of our biggest strategic challenges, our scale.

Selling Scale
First, why is scale a challenge for Sun? To be clear, I'm not talking about purchasing scale. As I've said before, we use innovation to drive product profitability, not simply bulk purchasing leverage. The scale to which I'm referring is selling and marketing scale. With Sun's current products, we could be selling to twice the number of customers we currently serve - our products appeal to an audience far greater than our customer base. But we're limited by our size - our sales and partner force has a tenth the resources of our biggest peers.

This is a particularly tough problem to solve in the midst of an economic downturn. Growing customers while reducing employees is an obvious challenge.

But it's also a huge opportunity. We have fewer than 100,000 customers worldwide. Using just one example, there are more than 10,000,000 MySQL users globally - reaching an additional 1% of them could more than double our customer base. The question is obviously how - we know we're relevant to those users, but we and our partners can't very well put sales reps on airplanes to visit all 10,000,000.

To answer that question, I'd like to examine what may seem like a tangential topic... the search business.

Discovering Intent
Now, why is the search business so valuable? Because it's an exceptionally efficient means of harvesting intentionality - if a consumer is searching for "flights to Cairo," the odds are good she's in the market for a trip to Egypt. That intent represents a ton of value for the airlines, hotel chains and car rental companies that serve travelers to Egypt. Whoever first recognizes that intent can broker a relationship between the traveler and those businesses, and charge a healthy toll for the privilege (that's the heart of on-line advertising). A discount airfare to Cairo, presented alongside the results of a "flights to Cairo" search, has a far higher likelihood of generating a ticket purchase than an unqualified billboard or ad in a newspaper. It's easier to find needles in haystacks when the haystacks are sorted by needle count.

Now I want you to think about the model I've described in these last few entries - Sun's business starts with exceptionally high volume free software adoption, literally millions of assets each day. What does that have to do with search?

Well, what is a customer telling us when they download software? Depending upon what they're downloading, they're telling us about what they value. If you're downloading MySQL or ZFS, you're more than likely storing data. If you're downloading, you're likely to create, save and maybe print documents. If you download VirtualBox, our virtualization software, you're telling us you work with multiple operating systems. An enormous stream of this kind of data funnels into Sun every day - signaling intent from customers spanning every corner of the world's technology market. That's the foundation of our analytical marketing activities.

Individuals and organizations opt-in to tell Sun, by what they download, what they're intending to do - which gives Sun a unique vantage point surrounding what comes next. If your company is downloading Lustre, the leading parallel file system for supercomputing, the odds are good you're on your way to building a supercomputing facility. Sun uniquely optimizes our solutions around Lustre, and we target those offers to an obviously interested user community. This is one reason we've been growing in the supercomputing market. We use software innovation to drive preference for Lustre - the majority of top supercomputing sites now use it. We target our product and service development to optimize for facilities using Lustre. And we target our selling and marketing activities around users that identify themselves to us - by downloading Lustre, or whitepapers and content related to it.

But as I've said, the majority of free software users aren't going to be building million dollar supercomputers, nor will they be issuing million dollar software purchase orders. And therein lies a new opportunity - one that helps us address our scaling challenges, as well.

Introducing Sun's Cloud
That opportunity is for Sun's Cloud - which we just announced today - to deliver commercial network services to the entire free software community.

Let's start with what we announced today.

This morning, Dave Douglas, the SVP of our Cloud Computing business, announced we're building the Sun Cloud, atop open source platforms - from ZFS and Crossbow, to MySQL and Glassfish. With more than 4,000 developers hard at work on these enabling elements, and a twenty year history of network scale software innovation, we're very comfortable with our technology lead. By building on open source, we're also able to radically reduce our costs by avoiding proprietary storage and networking products.

Second, we announced the API's and file formats for Sun's Cloud will all be open, delivered under a Creative Commons License. That means developers can freely stitch our and their cloud services into mass market products, without fear of lock-in or litigation from the emerging proprietary cloud vendors.

Third, unlike our peers, we also announced our cloud will be available for deployment behind corporate firewalls - that we'll commercialize our public cloud by instantiating it in private datacenters for those customers who can't, due to regulation, security or business constraints, use a public cloud. We recognize that workloads subject to fiduciary duty or regulatory scrutiny won't move to public clouds - if you can't move to the cloud, we'll move the cloud to you.

The Developing Cloud
How will developers use the cloud? Let me give you a very basic example - inside Sun, we're just now rolling out a version of OpenOffice extended for the cloud. If you take a look at the File menu in this picture, you'll see menu items that don't exist in your version -

but will exist in Sun's distribution. "Save to Cloud," and "Open From Cloud..." will enable OpenOffice users to use our public cloud to store and retrieve documents from the network, rather than their PC. We're in beta deployment inside Sun as we speak, and with around 3,000,000 new users joining the OpenOffice community every week, the opportunity to deliver this as a public service, to nearly 200,000,000 users, adn their employers, is really exciting.

The same applies to, say, VirtualBox - our desktop virtualization product, used by millions of users across the world. VB users will see a new feature later this year, offering an upload service to those wishing to archive or run multiple OS/application stacks - in Sun's Cloud. Those users have already told us they run multiple OS's - now that we know their intent, delivering a cloud to add value is a simple step forward. The same will apply to Glassfish and NetBeans, whose adoption helps us discover and recruit application developers - who might have a similar interest in running and/or storing apps in the cloud.

So in addition to offering the basic infrastructure services developers have come to expect (storage, compute, bandwidth), we'll be bringing tens of millions of free software users a library of cloud services and design patterns - designed to enhance the value they derive from the underlying software, while encouraging community development around open clouds. And all this will be based on what users have already told us they're interested in.

The Network is the Computer
To me, this is the embodiment of Sun's vision statement, the Network is the Computer. The breadth and quality of Sun's open source software is well known, and has created a user community that numbers in the hundreds of millions across the globe. The evolution of Sun's cloud and cloud services, from remote storage to remote execution, will allow us to grow our market, and the value we deliver to customers - even in, and perhaps amplified by the economic downturn. Clouds are just as interesting to students and startups as they are to Fortune 500 customers. If you're interested in Sun's Cloud, just head over to

The network is the computer has always been one of the most powerful statements describing the future of the technology we build. For the first time, we expect to translate that mission statement to our business model, investing in the free software community to grow our market, and leveraging the network to grow the value we deliver - to a market, and partner community, far larger than Sun.

And in that connection between adoption and commercial opportunity, we see near limitless opportunity, measured only by the scale of adoption we can achieve in a world where bandwidth is as pervasive as electricity, and free software adoption continues to accelerate.

With that said, this brings to a close this discussion of who Sun is, and where we're headed. I hope it's been useful. We're a very simple business, we strive to do three basic things. To drive free software across the world, both because it's good for the planet and innovation, and it's good for our business. Second, to deliver the world's most compelling technologies to captivate developers and deployers, alike. And finally, to put those assets to work in creating opportunities in the cloud, for our customers, our partners and for Sun, as well.

Thank you for your time and attention, I'll see you next time.

[Jonathan Schwartz's Blog]