Courtesy of Nigel Thomas comes this story of an IQ Trainwreck that eminates originally from the hallowed halls of Harvard University, where monkeys swing freely through the Ivy (original story previously published at www.thedailywtf.com as “I’ve Got the Monkey Now”).

The Harvard Business School Publishing redeveloped and relaunched their website back in 1999. Part of the QA was a series of test packs for their various search scenarios, including a test for ‘single result returned’. Being concientious people, the Harvardians regularly ran tests on the live system using the test packs to verify that the system was running correctly. Their single result test case was expected to return just one result.. for the search term “monkey” it was to return one article called “Who’s got the monkey?”. They put it in the shopping basket and completed a dummy sale to a dummy account (Mr 123, 123 one hundred and twenty third street… that kind of thing)

One day, in 2002, that test failed. An update to the paper “Who’s got the monkey?” had been commissioned and published called “Who’s got the monkey now?”. Why had that been commissioned?

 Well, it seems that the folks in the marketing department were looking at sales trends. How many of what was bought when? That kind of thing. Basic Business Intelligence.

“Hey Mike”, said a fictionalised HBSP marketing analyst, “there’s a lot of monkeys being sold on the site. Maybe we should upgrade the monkey”. “Yes Jim”, said Mike (who we must imagine as being a pipe smoker with a moustache for the necessary air of gravitas to be tangible), “that’s a lot of monkeys. Time to activate Marketing plan Ultra-Overdrive A!” (that conversation probably never happened like that, but in my head it made me laugh, so I thought I’d share it).

They produced summaries of the paper, pamphlets on the paper, knicknacks with monkey themes, widely publicised the success of the paper, and commissioned an update.

Which one day in 2002 wound up appearing in the single result search test cases run by the website team, causing the test to fail.

Why is this a trainwreck?

Well, Harvard spent a lot of money publicising a paper as a best seller that had reached that point most likely through test data being read as ‘valid’ when taken out of context. The ‘dummy’ addresses and dummy accounts could have (and should have) been excluded from the reports being used by the Marketing Analysts.

As an aside… I regularly get snail mail marketing from Harvard Business School addressed to Daragh O. Brien (note – part of my surname has become an initial).  Those of you who read the trainwreck about direct mail and direct email I posted last week can probably guess what happens to those letters.