4 ms·

The maturation of this technology has additional implications even in the absence of an open social graph. Imagine sophisticated FRT used to track individual cu

The maturation of this technology has additional implications even in the absence of an open social graph. Imagine sophisticated FRT used to track individual customer activity at retail locations. Timestamps could be used to map customers to their precise purchasing behaviors with infinite granularity. It's not far out to imagine a world where average profitability per visit is calculated upon arrival, now imagine an employee with a bug in her ear, marking you down as the guy with a 3 year history of strolling into Best Buy to browse TVs with the saleswomen, but never actually buying anything.

delinka 15y ago
"Timestamps could be used to map customers to their precise purchasing behaviors with infinite granularity." Between debit cards and loyalty cards, this is pretty much already happening. In the US, I see few people using cash without the loyalty card (because the loyalty card gets you a discount.) Of course, the remainder of your comment would require this new technology. And would be creepy as hell.
- vectorpush 15y ago
  The granularity I'm referring to is in the nuanced behavior of an individual in the act of shopping. Apparently, the algorithm has established that you follow a pattern of contemplating Cadburry chocolate bar purchases on every visit, well now your Inbox contains a buy 2 get 1 Free Cadburry coupon that only works for you. Once this is cheap enough that anyone with a high def webcam can consistently track/identify faces, this technology is going to be used by everyone.
- dhyasama 15y ago
  The difference is you could be identified when you arrive versus when you are checking out. Your experience within the store would change. For example, if you spend a lot of money then you might be greeted with VIP welcome wagon. If you don't, you might be greeted with a cold shoulder.
  - mike-cardwell 15y ago
    You could do this with existing technology. Just stick rfid tags in loyalty cards and a scanner on the door.
learc83 15y ago
This has major implications for loss prevention as well. External theft counts for about .5% of total revenue for the average retailer, which translates directly to a large percent of profit lost since margins are so low for many of them. There is a lot of money to be made for someone who perfects a machine learning technology to detect shoplifters.
- paganel 15y ago
  > There is a lot of money to be made for someone who perfects a machine learning technology to detect shoplifters This whole line looks like having been taken out of a Philip K. Dick novel. You're right, of course, which makes it even more creepy.

delinka 15y ago

"Timestamps could be used to map customers to their precise purchasing behaviors with infinite granularity." Between debit cards and loyalty cards, this is pretty much already happening. In the US, I see few people using cash without the loyalty card (because the loyalty card gets you a discount.) Of course, the remainder of your comment would require this new technology. And would be creepy as hell.

vectorpush 15y ago
The granularity I'm referring to is in the nuanced behavior of an individual in the act of shopping. Apparently, the algorithm has established that you follow a pattern of contemplating Cadburry chocolate bar purchases on every visit, well now your Inbox contains a buy 2 get 1 Free Cadburry coupon that only works for you. Once this is cheap enough that anyone with a high def webcam can consistently track/identify faces, this technology is going to be used by everyone.
dhyasama 15y ago
The difference is you could be identified when you arrive versus when you are checking out. Your experience within the store would change. For example, if you spend a lot of money then you might be greeted with VIP welcome wagon. If you don't, you might be greeted with a cold shoulder.
- mike-cardwell 15y ago
  You could do this with existing technology. Just stick rfid tags in loyalty cards and a scanner on the door.

learc83 15y ago

This has major implications for loss prevention as well. External theft counts for about .5% of total revenue for the average retailer, which translates directly to a large percent of profit lost since margins are so low for many of them. There is a lot of money to be made for someone who perfects a machine learning technology to detect shoplifters.

paganel 15y ago
> There is a lot of money to be made for someone who perfects a machine learning technology to detect shoplifters This whole line looks like having been taken out of a Philip K. Dick novel. You're right, of course, which makes it even more creepy.