IT Directions

Recently my wife has been trying to pay for the milk that, like every other child in her class, our daughter is given during the morning break at school. Parents have to settle the charge for this before the end of the previous term.

A simple enough matter, surely - a couple of clicks on a Web site should sort it out. Yet (so far) making this payment has required a series of telephone calls and dealings with four separate individuals at the organization who supplies the milk to schools, since the computer system that handles the payment was "acting up" in some unspecified way.

I wouldn't mention this if it was a one-off - an exception far from the rule. Yet a similar Kafkaesque scenario ensued when I tried to buy my son a new trombone. Arranging for him to use the school bus is a saga that will run and run. And after a recent experience with the UK government's online VAT service I may need counselling.

Is it just us? From the tales told by everyone else we know, ranging from health care to legal proceedings, I don't think so. The take-up of automation in our society is like the Emperor's take-up of new clothes - we're all just pretending it works.

I have learned the hard way to distrust any information generated by a computer. I have also started questioning the value of automation per se - especially of process automation. The interactions above would have been simpler, quicker and cheaper for everyone concerned before the organizations concerned installed workflow/BPM systems.

Pareto's rule tell us that the 20% of exceptional cases that require human attention result in 80% of the costs. Personally I would guess that the number of such cases is higher than 20% - and other commentators put it far higher. For instance, the Chief Strategy Officer of a major BPM company estimates that exceptional cases are 50% of all cases. If this is true, is the automated process helping or hindering people's efforts to get work done?

TAKE AWAY

Since first writing about Human Interaction Management, I have positioned the associated technology as complementary to mainstream BPM tools. I have always said that you need two forms of process software, in order to provide support for both human-driven and mechanistic processes. I am now starting to wonder about the efficiency of the latter - especially given the continually decreasing cost of human labour, and the desperate need for work among people in developing countries.

Why are corporations giving their millions to BPM/workflow vendors, when they could be building a committed and flexible workforce to deal with routine work by hand?

Is it an outlandish thought, that people might actually be better than machines at routine work?