% fortune -ae paul murphy

Sun's middle management

That blog entry by Jonathan Schwartz I mentioned yesterday used an anecdote about his experiences at two different hotels to illustrate the importance of having company evangelists on the front lines of customer contact. That story drew, among others, this response from "anonymous coward":

Unfortunately, those of us who deal with Sun bureaucracy and decision makers usually feel like we're staying at the wrong hotel.

Another contributor, "Will" had a longer version of this that's worth looking at carefully:

Sniff, Sniff....that was very touching. No other company in the industry, save for IBM (maybe), has the breadth of technology, services, expertise, hardware, and software (and free at that!!) you possess. My praise is real and beyond reproach ....but so is my criticism. Now if someone, perhaps you or one of your minions, could fix what's really wrong with Sun. The enormous amount of unnecessary red tape facing the reseller channel. It takes weeks for us to get pricing, and God forbid we forget to "dot an eye" somewhere along the way.....Well, then its back of the line with 'ya and the whole process starts over again. Its just so hard to get anything done, and frankly, sales people are predators. They seek to maximize their return (margins, customer satisfaction etc...) on products that burn the least amount of calories (cycles in computer speak). Margins are a whole separate rant....stay tuned. I've come to the realization that a shadow government, if you will, exists within the Sun empire. Mind you, it is not the individuals who work there, it is the system they've inherited. I defy anyone with any real power to change things to sit down with one of our sales reps, follow a fictitious (or real for that matter) sales opportunity through the maze (malaise would be a better word) that is the channel and look me in the eye and say "it is efficient". Bet you a dollar.......

To which all I can say, is me too guys, me too - a shadow empire indeed.

A few years ago I did some work for a western firm that had budgeted $23 million for a zSeries machine to run a large custom application being developed for them by an international IT services provider. That project had gone so far off the rails on both time and budget that they'd eventually taken it back in-house and staffed up to get it done themselves. Unfortunately that left them a year a late and something close to $16 million over budget. In response I made a number of suggestions, including hosting the thing on a pair of Sun 6800s instead of the mainframe.

On paper this not only got them out of the budget hole, but reduced expected operating cost by more than enough to accommodate re-development for the second and later releases using more "Unixee" tools than CICS/COBOL/DB2 and, of course, offered both improved performance and greater run-time flexibility.

It got a little heated, but the financial people eventually agreed to transfer about $19 million from capital spent for hardware to capital spent for software provided that the developers first demonstrate the application running reliably on Sun.

Unfortunately the company's size and U.S. parent made it a Sun national account - which meant we had to deal with Sun Canada's Toronto head office people and ended up undertaking significantly more software conversion work so we could run on HP gear - because having to fight the supplier simply for the right to buy from them didn't seem like an auspicious way to start a long term relationship.

So what's going on? There are many parts to this problem - starting with recruiting and compensation - but most are, I think, ultimately part of the expression of two main forces:

  1. first there's the loonovitch effect: the need to maintain quarterly results imposes a short term focus that in turn makes it very difficult for the company to nurture longer term opportunities -and whether that means supporting enthusiast employees or getting a customer from outside one's geopolitical comfort zone time in a test center lab doesn't make much difference; and,

  2. there's the tort effect: the need to avoid both real and imagined causes of civil or regulatory action can easily come to dominate formal decision making in U.S. chartered public companies.

To understand the loonovitch effect, imagine that you've just become a regional sales manager for Sun, you've got a short term sales deadline, a tough quota to meet, enough short term opportunities to eat up all of your available sales resources, and an empty sales position you're interviewing for.

Candidate A has a start tomorrow, short term, grab-the-money focus; candidate B, a work-with-the-customer focus aimed at building long term relationships. Which one do you hire?

To understand the tort effect imagine that your job calls for you to do, and say, things that are, or could become, public; but you know that everything that becomes public will be examined by hundreds of potential litigants and regulators looking for an angle from which to attack you and your company -for money, for revenge, for competitive advantage, or sometimes simply because they're on lifetime missions to redress social balance by gutting any corporate initiative they can.

To understand what's wrong at Sun, understand that the company has great R&D, great field staff, and an entreprenuerial, customer focused, executive suite -but middle management is entrenched, loonovitch driven, billings oriented, still smarting from customer defaults after the dot dumb collapse, and totally CYA focused.

Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.