Relaxing Constraints, or, Why We'll Win

I spent a day in Toronto last week and met with about 60 of our engineers, sales people, managers and marketing folks who work for Sun's software division. This was my second fiscal new year kick-off, having completed a similar event in Boston last week. I came home in a great mood, despite (a) the flight being an hour late (b) zero wireless service in the hotel or Pearson airport and (c) forgetting to eat dinner.

Why the happiness without sustenance, carbon, silicon or ether based? Quite simply, people are excited again. A bit cautious, a bit wary, but under the surface there's something brewing. I think happiness comes from relaxation - not just the personal type, but the corporate kind. We've relaxed some of the priority setting and organizational constraints that made us confused. Nothing like clarity of mission to energize you. Charles Garfield, author of "Peak Performers," described working on the Apollo 11 engineering team pretty succinctly -- they'd go outside, look at the moon, say "We're going there," then get back to work. Clarity is a winner.

Constraint #1: Make Money vs Grow. The first three words of the 11-word Schwartz Koan are "Make Money" and "Grow". These are great things to do if you're a public company. They're just really hard to do at the same time -- when you're focusing on returning to profitability, you prioritize spending, make painful cuts, and trim where you can. It is, in some cases, anti-growth, because growth requires investment. Once you figure out how to make money, you can spend some of it on growth. So the first big, hairy constraint out of the way is that our company figured out how to be profitable, and set goals this year for growth.

Constraint #2: Distributions vs Revenue Products. I've argued that the big change in software economics is the separation of distribution and compensation in the software world. Open source projects give us distributions; adding value provides compensation. In the last few weeks, we've organized software at Sun to deliver bits in these models -- Java Enterprise System and its Suites that are sold as products, as well as the large-bore distributions of open source projects like OpenSolaris, GlassFish, an implementation of an application server, and most recently, the OpenSSO code for extending and integrating our identity management stack. We're geared up to both distribute code as well as generate revenue from value-add represented in our code, whether it's a software product or a service. Distributions and products are co-dependent; there's no hierarchy or relative merit to one or the other. We have metrics for determining value -- we can count dollars, or users, or contributors, or downloads, or even metadata like OpenSolaris tagged items showing up in Technorati.

Constraint #3: Acquisition vs Strategy. Sun has received some deserved criticism for acquisitions that didn't always run according to the desired playbook. I'll argue that many, of those acquisitions were the result of trying to acquire a strategy, rather than having a strategy and acquiring the elements for its execution. We had no x86 or x64 strategy several years ago, but acquired Cobalt. Today we have a network systems strategy focused on assembling systems (and adding value through that assembly, something we've been doing for more than two decades). Acquired Kealia, got Andy Bechtolsheim back to execute on the strategy. We have a Microsoft interoperability strategy, needed to fill in some remote display pieces - Tarantella. We have a services-oriented architecture strategy, but needed tools to make it viable - SeeBeyond. The jigsaw puzzle solves when you buy the missing pieces, compared to buying pieces and then wondering what the puzzle was supposed to look like.

So where's the moon? It's not one huge bright target, but billions of smaller ones -- cell phones, PDAs, sensors, embedded computing devices, game consoles, PCs, desktops, and servers, a veritable Oort clout of networked people and things orbiting the Sun. Those users and devices are going to consume, and create, and share data. We're going there.


Hal, Great reading. E-mail me. I wanna catch up and chat. Steve

Posted by Stephen Beck on August 15, 2005 at 11:08 AM EDT #

Post a Comment:
  • HTML Syntax: NOT allowed

Hal Stern's thoughts on software, services, cloud computing, security, privacy, and data management


« July 2016