Rather than make the trains run on time, it may be better to do away the trains altogetherI acutally think I read an interview with Steve Balmer of Microsoft where he said something like that, but I don't think he took his own advice.
In any event, to give my friend a concrete example, I told him about a study I did on "the cost of a memo", circa mid-late '80's that goes like this:
- In the beginning, there was no personal computer. As a director, I wrote memo's to my staff in long hand (I hear that is no longer taught in school) and gave the scratch to my personal secretary to type on a Selectric with correcting tape. She then gave me the typed copy to edit, retyped it for my signature, and then she printed and distributed the memo through the office mail. I calculated the cost of a one page memo at about $50. (I may have under calculated, upon reflection, but this was '85 or thereabouts, so we worked cheap then)
- Sometime in the mid to late 80's, I bought my secretary a Mac and printer to replace the typewriter, but otherwise the process was the same. Cost went up to $60 to pay for the hardware.
- Then, I bought myself a Mac and I could type the memo and send it to her for editing and printing. The process saved a few minutes of my time but we had an extra computer, so the cost of a memo is now about $70. A lot of IT spending so far, no productivity gain, and a greater expense. General manager is not happy!
- Then came email networking adoption, and I could mail the memo myself. The cost has not gone down, still $70, but now the secretary has idle time (charged to the cost of a memo since the charges have to go somewhere). You can't fire part of a person, so we had to find something else for her to do.
- Then, memo's went out of style, replaced by the world of casual email correspondence. My time dropped a lot (per memo/email), so the cost per email dropped to about half or less of the formal memo, call it $30.
- Then, finally, with nothing to do, replaced by calendar applications and email, we fired the secretary, and the cost dropped again to about $10.
This arc took about 8 years to execute as we learned that with technology the idea is not to replace item for item in the process; it's to ditch the process and do something different with the technology.
I call this distructive improvement, but others talk about distructive innovation.
And we learned this lesson: putting in a bunch of IT personal application tools that save 10min here and there may not result in any real productivity or cost savings. It's rare that all these little segments coherently add to one headcount that can be released. You can't fire or transfer an FTE, only an integer person. So, as we saw in the 80's, and to some degree the cycle is beginning again, especially with ERP installs, there is capital spending without commensurate return to the business.
Another example, not project management, and further back: the transistor originally just replaced the triode tube, but the circuit design and functionality were not that much different. It wasn't until we grasped the idea that the transistor was a lot more than a 3 pin replacement for 3 pin tube did the whole digital productivity thing take off.
In any event: remember the wise words: It's often better to do away with the trains than to spend the effort to make them run on time!
