Technology Adoption (2 of 4)

As I referenced in my prior entry, I’m reviewing Sun’s three major strategic imperatives, and our progress going in to next fiscal year. Our strategic imperatives, in order, are:

1. Technology Adoption
2. Commercial Innovation
3. Efficiently Connecting 1. and 2.

This entry focuses on the first, Technology Adoption. Adoption is a non-economic phenomena, no money is spent, only time – yet it has extreme financial consequences. Let me give you an example.

I was with a big customer of ours last year, and reading through my account briefing before the meeting, I knew we were doing well. An analysis of their download activity showed they were heavy users of Solaris and OpenSolaris, and they had a large internal community of MySQL users, as well. In the meeting, their CIO said “we love where Solaris is headed.” I then asked if we could help with MySQL, and he said… “I banned it.”

Not exactly a buying signal.

I was stunned. I asked, “why?” He responded, “Oracle is our global standard, and with 20,000 developers, people need to follow the rules.” I said we had a very good relationship with Oracle, and started talking about how fast Oracle runs against our new Open Storage products.

Until he interrupted me, “…but my ban failed.” What? “We hire lots of people out of college every year, and they all come in knowing MySQL. All my prototypes are written to MySQL, and now I have a big base of MySQL apps I don’t want to port, and a bunch of MySQL programmers I don’t want to retrain. So I’d like a commercial relationship.”

In a nutshell, that’s adoption in action. Change in IT isn’t just a top down phenomenon – it’s more often bottom up*.

Innovation vs. Reselling Innovation
What’s the cost of missing that adoption? For Sun to resell a 1-way x86 server running Microsoft’s Windows or Red Hat yields (at best) a 10% gross profit margin. Very few companies have the scale to survive on those margins. More to the point, when you resell someone else’s products, your customer relationship isn’t built with the CIO or technology directors, it’s built with their reverse auction web site. For technology companies, the same applies to reselling any product you don’t own – it’s impossible to differentiate with anything more than a cheaper price. A price your supplier can, and will always, undercut.

Alternatively, when a user picks our products – when they build their storage on ZFS, their network on Crossbow, or their application on MySQL, independent of whether they’ve paid, they’ve created an opportunity for Sun – going forward, there’s only upside. It’s called positive option value.

Not to dip into finance 101, when the net present value of a lifetime revenue cycle exceeds the value of a one time purchase, a product or service that initiates the payment stream is either freely distributed (if it has no marginal cost, like software), or subsidized (if it has a hard cost). That’s why you see so many free credit cards, free checking account, free mobile phones, free month’s rent, free social networking, etc. In the technology world, free is the new black.

Free Markets
That’s also why the internet’s most valuable brands are *all* free – Amazon, Google, EBay, Skype, Yahoo!, Facebook, Hi5, MySpace, Baidu, TenCent, etc. Those brands reach more and have greater affinity than just about any other consumer brands. And in the technology marketplace, Linux, Java, MySQL, Firefox, Apache, Eclipse, NetBeans,, OpenSolaris, the same applies – free is a universal price, requires no currency translation, and reaches the longest tail of the market.

Now, could Amazon charge you to shop? Could your bank charge you to open an account? Google charge you to search? Could Sun charge people to download MySQL or Sure, we could also destroy those brands in a matter of days. If you’re not free, by definition you miss serving those that can’t afford, or aren’t ready to pay – which means your audience is capped, or destroyed if your competition is already free.

Microsoft’s the only company I didn’t include in the above list – and although I consider them a stupendously great brand, they’re the only company that can really approximate free while making money on the distribution of their products. The fact is they’re bundled on almost every PC across the planet, and appear “free” to the users who use those PC’s – they’ve amassed immense power with their distribution, and few users believe they’re paying for Windows when they buy a personal computer.

Thus, to developers (Sun’s target market) with Windows PC’s, Microsoft’s product are, in effect, already free. (As an aside, notice Microsoft inexorably moving toward free distribution, too, to reach new users – at some point, you can’t bundle every product on every computer, it’d be like printing a Sunday edition of the newspaper every day of the week).

This is exactly why we freely distribute our key software assets all over the world – if we didn’t, users and developers might pick someone else’s free product (or simply use the one they assume to be free). And if they picked someone else’s product on which to build their business or their application, Sun becomes a reseller – which isn’t our mission or business model. It’s a free market, in every sense.

The customer I referenced in my first entry

that said, “I haven’t visited Sun in five years, but all of a sudden you seem to matter to my developers” was saying he was seeing exactly that, a lot more of our products used by his developers – from VirtualBox to MySQL, Glassfish to ZFS. For some users, and nearly all developers, budgets aren’t measured in dollars, they’re measured in time and attention – if you want those audiences to spend their time and attention with you, you have to earn it. If you earn it, a preference forms. For Sun, we drive that preference over our competition, primarily proprietary alternatives.

Our Products are Our Ads
Now, the words “driving preference” are used by the advertising industry when talking about branding. Businesses brand or advertise to drive awareness of or preference for their products. In the case of a Nike or Toyota, both have to spend fortunes to “buy media,” or acquire the ad space (or airtime) through which they’ll present, free of charge to consumers, the images or content they feel best represents their brands.

Why doesn’t Facebook advertise? Because Facebook itself is a branding experience. Using Facebook drives preference for Facebook. And their audience, in users, outreaches just about every media company on earth. It would make no sense for them to buy media, they are media.

For the audiences Sun cares about, those building, deploying or buying technology, we’ve got a similar reach. By being freely distributed, our products build their own audiences. And using the products, from Glassfish to ZFS or NetBeans, creates a branding experience (and a wildly positive one, if we’re doing our jobs well). So why don’t we advertise in traditional outlets? Well, every day, the number of people using our products, getting that positive branding experience, eclipses nearly all major newspapers globally, combined.

By proliferating Sun innovations, even encouraging derivatives that will never drive revenue to Sun, we are creating preference for open source, awareness of Sun as an innovator, and displacing proprietary vendors that can’t build comparable audiences. That preference has value to us, and to the broader communities in which we participate. The value spans awareness, market penetration, skills development, ecosystem expansion – a healthy community is a growing community.

How else is adoption or preference valuable? Volume adoption attracts application developers, and can drive tipping effects – once one independent software vendor, or ISV, picks your platform, others that work with that ISV follow suit. If you do a good job, you lead an avalanche of ISV’s to pick your platform – which makes it more appealing to end users. That’s why Red Hat has such a durable Linux model – once Oracle picked Red Hat as the Linux to which they’d first certify Oracle’s database, the ISV’s that relied on Oracle certified only to Red Hat, which tipped the market to Red Hat so strongly that not even Oracle has been able to undo that grip.

So adoption drives the ecosystem, which drives more adoption and more expansion… you get the idea. It’s a virtuous cycle, a cycle that starts with volume adoption.

What Adoption Looks Like
So what does “Adoption” look like? Here’s a picture, which shows the ramp we’re seeing from free software adoption across the world for some of our key datacenter assets.

For competitive reasons, I won’t specify which products are shown, but suffice it to say we’re very happy with the ramp, and we love great product reviews – that’s what drives the spikes. The troughs are weekends. Bear in mind these are datacenter assets (like Glassfish and OpenSolaris), not consumer runtimes (like Java or, so these downloads influence datacenter design. Every day, our software is working hard to drive preference in startups, in government agencies, dorm rooms, ISV’s, fortune 100 IT shops, everywhere the internet reaches. Free products reach everyone interested in them, no barriers.

On the consumer side,, which certainly promotes Sun’s vision of open standards and data formats, reaches nearly three million new users – every week. Adding them to a user base we estimate to be between 150 and 200 million users. Talk about global circulation.

Our products are our brands and one of our most effective means of driving design wins – in front of users, developers and OEM’s. In markets as diverse as high performance computing and grid scheduling, web databases, application infrastructure and desktop virtualization. Free distribution and access to source code is our investment in the global developer community. We invest with our code, our ideas and time, and we promote and encourage derivatives. We gain by reaching people we’d otherwise never reach – and earning their attention and engagement. Even if we’re never paid, that’s positive option value.

Another way of looking at adoption are these “pink dot” maps – they show us where our products are gaining users via opt-in registration. My favorite dots on the map shown are in places we clearly have no sales coverage – I’d like to say hello to our users on the Falkland Islands, thank you for choosing Solaris πŸ™‚

What’s the value of all that adoption? Like the value of search, shopping, or opening a bank account, there’s no instantaneous value beyond the fact you’ve chosen to invest your time and energy in our ecosystem, and not our competition’s. At global scale, that makes us an enormously tough competitor for proprietary companies, or those without true innovation. For example, to get a sense for what our proprietary storage competitors are facing everywhere around the globe, Google the phrase, “love ZFS“.

But as with all free business models, the real value arises in what comes after free – and that’s my teaser to get you to read the next blog entry, focused on our Commercial Innovations.

Thanks, again, for reading, watching and commenting.


* as I often say to groups of CIO’s, “which one of you gave permission to your employees to search on Google?” No one ever raises their hand πŸ™‚

And to my readers/viewers for whom English isn’t your first language… I’m doing my best to talk more slowly. I will redouble those efforts in the next video… thank you for watching!


Filed under General

26 responses to “Technology Adoption (2 of 4)

  1. To directly influence developer. Make glassfish / mysql available as google app engine. I use tomcat today just because I know it very well(implying is the incumbent) and switch to glassfish will is not easy (will potentially eat up 16 hrs for start time) and potentialyl more if I run into issue’s – which is quite probably

  2. I would consider adding a token $1 charge to all future downloads of mySQL and the like. That’s the only way you’ll be able to judge the true loyalty of the fans of these software objects. If people aren’t willing to contribute a measly dollar to these software objects, you have to ask, "why?" It’s only a dollar.
    Reagrding Oracle, I’m going to go see if I can pick up some schwag over at Redwood Shores. Maybe there’s some best practices that mySQL could adopt/integrate into her code.

  3. Kevin Heatwole

    Why don’t you release ZFS for Linux under GNU license? The version of ZFS available for Linux is not so good due to your license. I’d love to try it out on my Centos servers.

  4. Jon

    Well done. Nice post. I look forward to the next two in the series.
    To foment faster adoption, may I make a suggestion? It may be helpful to host a "Development Days" road show where you can set up an exhibit to display some of Sun’s technologies such OpenSolaris. I talked to some of your folks at the last JavaOne and they were VERY helpful and did their job to pique my interest. Figure out a way to reach out to the people who can’t attend the JavaOne-like big tent shows. Offer to have Sun to give presentations at various JUGs and LUGs, and even various universities.
    It seems you’re on the right track. Now reach out to more people in real life.

  5. I think you’re going to have some issues with Java adoption on Apple’s computers and devices. They don’t run your latest version of Java, and for me, Java doesn’t run in Firefox on my MacBook (although it seems to kinda work in Safari). Good luck with that.
    I’d also like to see you measure adoption more finely. You just seem to look at downloads and sometimes registrations, but you could also look at real usage numbers (especially with NetBeans). So the lifecycle would be:
    1) Download (measured)
    2) Install and register (measured)
    3) Use and use again (measured, e.g. average hourly usage per week)
    4) Install modules/components (measured)
    5) Buy services/subscriptions (measured)
    So rather than just hearing you say "wow OpenOffice was downloaded X times last week" I’d love to see the progress on the lifecycle steps from 1 through 5 above. And the same goes for NetBeans, OpenSolaris, MySQL, and all your other brand assets.
    Step 5 is still not intuitive or natural for your software downloads. There is no button inviting me to "save to the cloud" in OpenOffice, or "run on the cloud" in NetBeans – and these are revenue generators for Sun down the road. I’m sure you’re working on it.

  6. Nathan Evans

    Thank you Johnathan for giving us to use such great products.
    My Lenovo t500 laptop which is almost 100% supported by Nevada b108(wireless was supported out of the box I was happy) sports your logo. I carry it with me everywhere I go. People have asked me "Whats that?". I always repsod the same way. "This is Solaris, the most advanced operating system on the face of the earth." I enjoy telling people about Solaris, Java, and Sun. I think you went the right way with your current business model. I am currently in college and already several of my peers have asked for a copy of Nevada.
    I also think it’s great that I get to "interact" with the CEO of Sun Microsystems via this blog, it’s great send me an e-mail sometime.
    Oh, I only have one favour to ask, can I have a e-mail address
    or maybe a Sun sticker for my laptop??

  7. Really good post. This is clever. You are the first and only CEO I know, that talks openly about his company to everyone, employees and customers.

  8. Steve

    I’d second Kevin’s comments about the download verses actual longer-term usage. For example, I and my colleagues downloaded Open Office some months ago onto our Windows desktops with the intention of evaluating it, but haven’t looked at it since and continued to use MS Office. Did these downloads count towards your adoption success figures?
    I’d love to see some details of how, for example, Glassfish is being used in real-World scenarios since download figures really don’t mean an awful lot to true adoption within a company.

  9. Paul Fookman

    @Mark, if Sun charged $1 for MySQL, not only would I stop using it, it would removed from every Linux distribution on earth, and you’d see a quiet, and efficient, migration away from it. It would take about three days. Read the blog, he’s got his head on straight.

  10. Stephen

    I have spent much more money on hardware, books and services since I switch to open source software. There is money with this business model. I use Open Office and have been interested in trying Open Solaris. This guy is right. Once you use their software you can then be sold additional services and add ons.

  11. Mike

    re: "For some users, and nearly all developers, budgets aren’t measured in dollars, they’re measured in time and attention…"
    You’ve made a powerful statement acknowledging that superior customer service makes a difference to your customers, regardless of the economic model you’ve chosen for delivering your products.
    You also give the example of MySQL devotees driving status change from a banned product to a preferred database platform. Unfortunately, not all of your units understand the power that their customers have to drive change to, or away from, your products.
    A three month old VirtualBox support thread dealing with regressed disk space management features has a number of people bugged. We’ve lost the ability to shrink our storage volumes for the past two versions. Please see the support ticket at
    The same people that you admit drive adoption of technology are rewarded with comments such as "You are in no position to tell us what to do or not. No amount of complaints from your side will change the priority of this issue."
    Well, maybe or maybe not. Is the person who received this answer one of those CIOs or perhaps an influential developer who is in a position to adopt your technology or not?
    Maybe we can get a little more of your "time and attention" filtering down to the VirtualBox support group? Or at least have someone give them a week in charm school?

  12. benr

    The point about "effectively free" is a good one. Solaris was in this camp for a long time. Even though some people (competitors) would refer to Solaris as "expensive and proprietary", there were no Solaris serial keys and as far back as Solaris 2.6 you could freely download the ISO images.. the cost was baked into your hardware purchase and it was, therefore, effectively free.

  13. to Mike

    oof. that is harsh, inappropriate and plainly NOT ‘superior customer service’. In fact, it clearly IS how greatly the impact is felt that should drive issues to the top of the list. Who is this "sandervl73" and who made him the decider? And what of "techtonik"s question? Can a workaround be built to give some relief in the meantime? Hopefully this will have a happy ending.

  14. It is interesting story. I use mysql for my site.

  15. Santhosh

    @Steve, provides some detail on how Glassfish is being used by a variety of businesses. lists stories broken down by Sun’s product lines, including software.

  16. Lee Hepler

    About downloads versus usage. I downloaded one copy of Solaris at home and installed it on all three of my Sparc based computers. I downloaded another copy at home and installed it on one X86 workstation. I downloaded another copy at work and installed it on 50 Sparc based computers that belong to the government. I agree that downloads do not represent a one to one correlation with installs but this may be a one download to numerous installs reality.
    I worked in computer security for three years and had over 220,000 computers to oversee. One person would have downloaded Solaris once for over 10,000 computers in that situation because my office had to certify the software before installation and provide the OS and secure configuration instructions for every installation.

  17. James Mansion

    Another great post. However, it still leaves me very uneasy, for the state of our industry. If everything is ‘free’ for users to acquire and there is a huge temporal disconnect between R&D costs and the lifetime revenue, then only those with very deep products (or incredibly low R&D costs) will be able to play at all – and venture funding multiples will go very high too, since that temporal disconnect has a big correlation to risk of (any) return. I don’t think this is a good thing at all, and it strikes me that we haven’t learnt very much from the ‘market share at any cost’ scramble that was the first dot com bubble.
    And no, I don’t buy ‘incredibly low R&D costs’ as a solution – software is costly to develop, and relying on charitable donations from those with nothing better to do is risky, not least because the control aspect isn’t there and, as you say, then you’re just a reseller – or ‘distribution vendor’.
    Some of us, I think, will prefer a profitable niche that doesn’t involve betting with granny’s savings, either directly or with a venture manager taking a haircut. You at Sun are at least putting your corporate money where your mouth is – but let’s remember that your results *so far* aren’t lending huge credibility to the approach. Whether they *will* do, only time will tell. And only time will tell whether your shareholders are true believers with the necessary patience.
    It is, after all, their money you’re gambling with.
    There may be trouble ahead – so let’s face the music, and dance!

  18. Softwareman

    Jonathan, it is this creative, forward thinking and strategic look at driving technology adoption (and sales) that lured me to Sun many years ago. The irony is that after eleven years it was the phasing out of the customer facing Solaris x.86 organization that cost me my job with Sun. IBM and HP are utilizing Solaris on their DL systems and BladeCenters to catch customers coming doen stream off of SPARC. Who’s out there driving adoption of Solaris on those platforms with Sun market share and revenue in mind?

  19. Adoption time does matter a lot for a successful delivery of project, as bieng a project manager you have to meet timelines.
    One quick question, by influencing the man power this way, doesnt it makes it harder for a business analyist or solution architect to manage technology tradeoffs. A team might not be able deliver the optimum technology just because there is not enough man power available for it?

  20. Another great video. These videos are more than The State of Sun, but basically a how to guide on running a business, in particular one where free distribution of some portion of the company’s products are used. It really gets the gears turning.

  21. Why don’t you release ZFS for Linux under GNU license? The version of ZFS available for Linux is not so good due to your license. I’d love to try it out on my Centos servers. " thats true. thank u πŸ˜‰

  22. I like OpenSolaris a lot, but unfortunately it is inadequate for modern C++ development. While the SunPro 5.10 compiler is commendably strict and emits efficient code, the standard library is hopelessly behind times. To use Boost (a minimal requirement for professional C++ development IMHO), one has to use STLport V4 (which is also outdated) or has to install e.g. the Apache stdcxx first. Too many hoops to jump through.
    If you target the high-end server market, make sure you win the hearts of systems software developers.

  23. Thank u Jonathan. U have a very nice blog and important articles!

  24. Notmyre Alname

    I fully support Ben Snow’s statements about [Open]Solaris being a suboptimal C++ development platform. I just hit a SunPro CC bug recently, which is apparently present since the 5.3 release…(!) The bug number is 6757317, one gets an error "__SLIP.DELETER__F() called for const object" on a perfectly harmless virtual dtor… Should not happen, should have been fixed long ago.
    Clean up Sun CC. Do not let ‘gcc’ become the one and only C/C++ compiler on Solaris. This platform deserves something better than that.

  25. Enjoyed the videos and post. Very informative. Bottom up adoption feels like the new corporate culture.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s