The Ultimate Business Machine - Archives
List of Categories : Database * Technology * Commentary * Singapore * Travel *
Sun 25 May 2003
Category : Commentary/connections.txt
I'm trying to relate the idea of making a good living, through running a business that works like a machine, to knowing how to find the pieces of technology that can help you to actually build that business machine.
Without a system, a business cannot scale, and the owners cannot disengage from the daily grind long enough to dream about where to take that business next. With a system, you've already built the procedures that can be done by rote into the system. That forms the spine of the business.
If you have a good system in place, you can have a lot of flexibility in planning your staffing. You can experiment with the combinations - hardworking journeyman types, to people with a lot experience, to the young, fresh and idealistic - because a good 80% of the business processes have been made to run like clockwork.
In a system we have in place for a client (that's regrettably withdrawing from here, but had contributed a fair bit to our own passive income), they recruited a 65-year-old temp with vast claims-handling experience a year back. Old though he was, I've observed that it was possible for him to learn the system in a week - score a point for the Mac school of user-interface design.
He's helped by a system that presents him with only the data that needs to be focussed on, that collates the data and shunts it to the next step in the process, that knows how to roll everything back when he makes a mistake. That allows him to concentrate on what he is good at.
But I did not come to praise our system. No, only to bury the idea that IT has to be talked about in terms like "three or four-tier client-server", "hot or cold start", "Used-Case diagrams", "Black-box or White-box testing".
I believe there is a place for everything, and everything in its place. The trouble is that the IT Manager as GateKeeper is in the wrong place.
To build the Ultimate Business Machine, we've got to start the discussion from the business end of things, from the systems, procedures and processes that are needed, and work progressively downwards into the technical implementations, making innumerable trade-offs along the way. All the while, we need to remember that we're trying to build a "machine" that will deliver consistent services, of a consistent quality. We're going to have to resort to triage, making decisions in favour of cases that work 80% of the time, and move on.
I believe that if you proceed along these lines, you can force each technical decision to be reviewed against its contribution to the business objective - does it make the business run like a high performance engine? - and then choose the right tool for each job. It's possible to hold a technical discussion wholly at a descriptive level, in terms of capabilities. It's possible to insist that a technical person speak plain English. Get him to draw a picture of what he's trying to do. If he can't, it's possible that he's really clueless. You've got to have faith in your ability to discern the truth.