IT in the business management process needs a human touch and trial and error
Matt Senatore, VP Global Marketing Casewise
As part of an insightful morning at the Gartner EA Summit in London yesterday, I had the pleasure of listening to keynote speaker, the Undercover Economist, Tim Harford.
Introduced by Gartner’s Brian Burke as “being blessed with the talent of making the complex simple”, Tim engaged the audience with a series of quite unrelated stories which indirectly highlighted two key considerations for any IT business process management – trial and error and human intervention
If at first you don’t succeed…
Tim initially recounted a seemingly bizarre story of a student who wished to make, from scratch, their own version of a simple house hold item – the toaster. The inquisitive students’ approach was to acquire a cheap toaster and strip it down to see what parts it was made up from. Staggeringly, he discovered a single toaster comprised over four hundred individual components parts.
The student laboured for months to create a basic toaster, involving use of a microwave to act as a smelt for the raw Iron Ore (I’d imagine this highly dangerous process took more than one go) through to using a tree trunk as a moulding for the plastic outer casing.
Finally ready nine months later, the completed toaster lasted approximately five seconds before bursting into flames. As Tim explained, this demonstrates the multiple issues you need to address when you want to achieve anything – even the build of a simple toaster became such a complex, as well as the stages of trial and error you must endure.
Tim further built on this theme describing Unilever’s creation of a simple pouring nozzle for one of its detergent products. The methodology adopted by the consumer goods giant was to create ten different nozzles, then find which one worked best. From the best performing pourer, ten more modifications of a similar design were produced and the process repeated until the very best result was found. To this day no one actually knows how or why it works so well, but it just does – through human input combined with trial and error.
With these two verbal analogies, Harford described how organizations faced with changes – no matter how simple they may appear – need to use trial and error in their processes in order to achieve their objectives.
Right here, right now…
Harford went on to explain how IT, though key in the business management and Enterprise Architecture environment, is no substitute for the “there and now” knowledge of the person in any situation.
Sharing an exemplar tale from the first Gulf war, Harford regaled a story which could have had distressing consequences, if human intervention had not played a role…
Following orders, issued based on key information from a central computer, a US Army tank division raced through the desert in a terrible sand storm. As the unit cleared the top of a ridge they realized they we driving straight into a platoon of Iraqi troops hiding in a dug out just beyond their position. According to the data feed being received by the tank division, the enemy army didn’t exist at that point – though clearly the men on the ground could see a different scenario and needed to take immediate action.
In a critical, split second decision, the General leading the troop ordered them to attack. Within twenty five seconds the superior US machinery, fire power and trained troops disarmed the enemy unit. Meanwhile, back base there still wasn’t clarity whether Iraq were actually there or not.
In sharing these simple yet effective examples, Enterprise Architects and Business Process Management Specialists can learn the following:
• Start by finding out what you need to know
• Be prepared to underestimate a problem the first time round
• Keep testing and learning, repeatedly
• Listen to your men and women on the ground, respect the knowledge of your subject matter experts and build this into your process
And, finally, yes you will make mistakes, but try make them good ones!
So, my question is: what good mistakes have you made that have improved your processes for the better, in the longer term?



