The future of AI in marketing
May 23rd, 2019
If you follow the KDR blog you’ll know we love writing about big data and how it is being used in the real world, whether this is helping beat poverty, preventing fraud or even just making sport more entertaining.
But it is Halloween, so it seems like no better time to take a look at the dark side of data and what happens when the data goes bad!
Earlier this year a man’s house in Texas was torn down because of a Google Maps GPS error. The demolition workers had relied on Google Maps to take them to the correct address, however Google had the wrong location and the real house was actually a block away.
And what’s really scary is that this isn’t the first time this has happened in Texas!
This one is straight from Hollywood. During the making of Toy Story 2 one of the employees accidentally pressed the remove all button, when they went to the backup drive it turned out it hadn’t been working in months.
Luckily the Technical Director had made a personal backup drive!
One beauty site allowed users to have their email address as their registration with no password, this meant that when someone wanted to check out quickly they could put their email address in and all their personal details would appear.
Seems fine, until someone enters other people’s email addresses and gets their personal details as well.
Data Loss (again)
In 2015 many drivers across the country found themselves having their driving license taken off them for no reason and having to retake their tests. One driver sent his license to the DVLA to update his address and when it returned it was missing his entitlement to ride a motorbike and when he phoned to complain he was told he could lose his full driving license as there was no record of him being able to drive!
It turns out the DVLA’s systems were deleting driver’s records.
Check out IQTrainwrecks.info, run by data expert Daragh O’Brien, they gave us plenty of inspiration for writing this blog and have some real data quality train wrecks!
These are our top data horror stories but there are plenty more out there, and sometimes it’s not just major data fails that scare us, as KDR’s Leon said “receiving a mass-targeted email with the first line as “Hello @FIRSTNAME” always makes me cringe!”
What are your data horror stories? Do you have any first-hand experiences with “bad data”?