There are only two ways into the technology market, the front door and the back door. Some examples that came in through the front door: ERP, mainframes, and Lotus Notes. Back-door arrivals: personal computers, Unix, and Dynamic Languages (Perl, Python, and so on). You can build a business both ways. And, now that I’ve been here at Sun for almost six months, I wonder: Which door should we be knocking on?
The Back Door and the Matrix · Last year, I published the Technology Predictor Success Matrix, a methodology for predicting whether a new technology will go over the top. When I say “back-door” technology I mean what the matrix called Happy Programmers. The idea is, if the back-room geeks like something, and find that it’s useful, they’ll start using it whether management thinks it’s a good idea or not.
The Front Door and the CIO · I’ve dealt with quite a few CIOs, and they don’t think about technology, they think in business terms and those are the only terms they think in. When the SAP salesman says a company can cut acquisition and logistics costs, or the IBM salesman says Notes can get people working together better, nobody thinks much about what the architecture is or how it works or whether programmers like working with it. Both those salesmen came in through the front door.
Sun’s History · Sun was historically a back-door company. Management and the analysts and prognosticators and journalists were all looking at IBM’s operating systems and DEC’s VMS and a bunch of other now-forgotten software, and meanwhile, the geeks were loading Unix boxes in the back door because they got things done, and Sun made good ones.
Then, as the Nineties went along, there was a one-two-three punch that found Sun, dressed in a suit and tie, coming in the front door. First, more and more applications migrated to relational back-ends, which tended to run better on Unix. Second, the Web tidal wave hit, and most Web servers were Unix boxes, if only because TimBL had been sitting behind a desk at CERN that had a NeXT cube on it. Third, Java came along, and a wave of enterprise software developers ignored all the web-page-animation hype and noticed that this was a lot better than what they’d been using for big back-end applications.
This was a moment of rare harmony between geek and management world views. Geeks liked the Web, and management saw that FedEx could track packages without having to ship client software or pay someone to take phone calls. Geeks liked Java a lot, and management noticed that they weren’t being asked to make twenty-five year bets on a particular system vendor as a precondition of deploying an app. And there was Sun standing on the front steps, and to its credit did a pretty good job of learning to wear a suit and talk to CIOs about business issues.
Today · If you go to JavaOne, you find yourself among thousands of (mostly) enterprise developers and their bosses, and the three things they care about are stability, predictability, and reliability. The Java Community Process does a really good job of meeting these people’s needs. It isn’t lightweight, it isn’t real friendly to OSS hackers, and it isn’t particularly cheap, but if you talk to the people inside (I’m not one of them), most seem happy with its trade-offs.
Sun’s sales force is very front-door these days, their job is to deliver solid, reliable business infrastructure in support of business goals and that’s what they do and that’s why Sun’s been cash-flow-positive even as we and our markets shrank over the past few years.
But I gotta say, I don’t hear about much Sun stuff coming in the back door these days because the geeks like using it.
Where to Grow? · So, we’ve got pretty clear marching orders from the boss, and here I’m focusing on item #2, “Grow”. Growing the front-door business is a conventional (if hard) business problem: we have to build boxes that are winners on cost and performance and make them easy to buy/deploy/manage and figure out saner pricing schemes, especially for software, and take good care of the JCP, and structure our sales force to align with customers. This is what McNealy and Schwartz and Fowler and company are getting paid for and based on what I see, we should do OK.
Which brings us to back-door growth. There are two things we have to do here: first, start shipping more technology that that geeks like and use irrespective of this month’s management mantra. The second is just to become a little friendlier and more conversational. I’m sorry, a public company as big as Sun is never going to be hip or have serious street cred (I think there’s a whole division of the SEC dedicated to preventing hipness). But still, compared to shipping the right tech, being less uncool is the easy part: with Jonathan and Mary and Liu and Mortazavi and 500+ others blogging away furiously, at least we’re not boring any more. And we’re listening a lot better.
Building Back-door Tech · First off, there’s no point trying to do corporate planning on this, you leave the geeks to themselves and it either happens or it doesn’t. Given the geek density at Sun, it probably will... In fact it is, DTrace is nerd fodder if I’ve ever seen it, and I think some of the hardware we’re building will qualify too.
But there are a few directions I think we should be looking, and they won’t surprise anyone who’s been reading ongoing.
First, we should get serious about dynamic languages. I totally don’t see any reason why we shouldn’t try to become a world leader in this space. They are the only thing on the horizon that’s going to steal mindshare from Java in the enterprise IT ecosystem, and geeks love ’em so we should lead not follow.
Second, we should be the first with the most on syndication technology. Rome is a step in the right direction, and there is some other less public stuff, but this will more than repay some short-term focus.
Third, there’s tremendous juice to be had around the edges of the WS-* swamp. While IBM and Microsoft play dueling committees and dueling specs, I observe that Google and EBay and Amazon and the like are doing tens of millions of real Web Services transactions a day and Sun should ship the plumbing to make it easy to build that kind of app and talk to that kind of app, what a no-brainer.
Then there’s my own Zeppelin skunkworks, if I can just dig out from under Atom and the other internal admin work, I think it’ll have geek appeal. I’m one, and I only know how to write software to scratch my own itches.
And that’s just the stuff I know about, and I haven’t even been here six months yet.
The Alternative · That would be to ignore the back door and decide that if CIO’s don’t want it today, our engineers shouldn’t be working on it. Wait, I’ve seen that movie before... it was at a computer company, second-largest in the world when I worked there, called Digital Equipment Corporation. R.I.P. That approach would not only be fatal, it would be boring. Let’s not do it.