Previously, previously, previously, previously, previously.
Not only that, but looking at the opt out option, you have to give them your MAC address!
Just wait. Once somebody scripts the entry of all possible MAC addresses, it'll require a CAPTCHA too.
By the way, anybody with ff:ff:ff:ff:ff:ff, I got your back. :)
What, no love for DEDBEF?
And for future reference, here is the opt out page.
I suggest people seeing these signs go in the store, find someone who looks in charge and explain why you're going to visit a different store now. Turning off your phone to avoid being tracked is crap.
Stupid ideas like this need killing off before they become normal.
Walking around with an actively broadcasting radio transmitter and then complaining about other people being able to receive your transmissions and thus get your approximate location is a bit silly, don't you think?
I suppose a "do not try to connect to any random wifi hotspot you find" option in phones might be sensible, since I don't think simply scanning for wifi networks involves broadcasting anything, so surely they can only track you when your device tries to connect to whatever wifi hotspot they're pretending to be.
Wifi doesn't obey walls and property boundaries though... so what about all those people merely walking past the place, or in the shop next door that get logged and tracked?
Thinking more though, how is the information they collect even remotely useful? Surely all it can collect is "people with smartphones spend on average X minutes in your shop". It's not like you're going to get GPS quality "your baked beans aisle is really popular" stats.
You think of that as a problem. They consider it a feature. Look at "walkbys": http://euclidanalytics.com/features
> Thinking more though, how is the information they collect even remotely useful?
"We need to know which part of our store gets the most foot traffic, so we can put our most marketed products there."
"We need to know which part of our store gets the least amount of foot traffic, so we can optimize the store's layout."
"When people walk out without waiting at the cash register first, what does their path look like? Do all those people have something in common?"
"On average, how long does a person wait in line before being able to check out? How many people give up while standing in line and leave?"
Lots of useful data to be gleaned from this, and this is just off the top of my head. I'm sure marketing people have thought long and hard about this.
Yeah, except it's wifi, not GPS. Their features PDF makes it just sound like the physical version of Google Analytics, talking about "bounce rate" "repeat customers" and so on.
It can't work out where you are inside the store, where you've walked or how long you were queueing at the till. All they collect is your MAC address and use it as an anonymous ID to count the amount of time you come in the store.
It just lets you work out "how many people with wifi enabled phones went in my store?" "how many of them are repeat visitors?" and other bland statistics that really... your tills and your staff can work out for themselves using good old human interactions.
Euclid claims to be able to determine your approximate position: http://euclidanalytics.com/features/product-more-detail
More detail on how it works, please.
The Euclid sensor measures Wi-Fi “pings” that are broadcast by Wi-Fi enabled phones as they search for available networks. Our advanced heuristics use several different factors – including signal strength, ping frequency, and proximity to other sensors (if any) – to determine the phone’s approximate location. The phone does not actually need to be connected to a Wi-Fi network; it merely needs to be open to the idea.
Plus, they list some of the useful things you can gather.
As far as using human interactions, you might as well argue that they ought to through away their tills and use abaci and paper ledgers instead. This might be a creepy invasion of privacy, but the Golden Age argument doesn't work particularly well. (Not to mention, using "good old human interactions" doesn't count how many people come into the store without buying anything or interacting with anyone, and it definitely doesn't count how many people stared at your window displays and then walked off.)
Not quite the point I was making.
You can work out how long shoppers wait at tills by simply asking the staff - "was the shop busy today? Were many people stood in queues?". Shops already do this if you listen to them tannoy for more till staff, etc.
Store cards exist solely to track what we buy and how often we buy it. As does tracking your credit card details.
I'm not sure their tracking system could tell the difference between "someone looking at a store display for five minutes" and "someone stood waiting for a bus" or "someone stood staring at their phone on Facebook". Their whole system relies on a side effect of how mobile phones work, rather than an actual feature.
You know, employees with their own phones will really skew their stats - "the staffroom is the most popular part of your shop, with an average visit time of six hours. You have 25 visitors that come every single day for 6 hours".
And this technology would explode if a shop next door to an Apple store tried to use it - those things are so active you can almost taste the wifi in the air :-)
The invasion of privacy bit is irrelevant, they're just capturing your MAC address, and the shop CCTV will be recording your every movement anyway (there's something to invent... HD quality CCTV with face recognition and person tracking). No, the problem is that this is a really crap idea that will be dead in a few years - either it won't catch on, or people will scan the QR code to opt out.
I think you're intentionally missing the point about how easy this data is to mine. Behavior of employees is easy to filter out, because after a few "break room" points you can trivially train your rule-set to know which MACs belong to employees.
As someone who has employees and does care about customer flow, I can promise you that asking them to open their eyes and observe what is going on around them is completely fruitless. But even if it wasn't, this is additional information on top of that that may complement or contradict what they (think they) observe.
Even if this system does not in fact accomplish what the hype says it does, systems that actually do track all of those things are pretty obviously only 20 minutes into the future. If it can't figure out precisely where someone is, that's just a bug. I am confident the next rev will fix it.
Basically any arguments against this kind of system are identical to arguments against tracking web clients with cookies. Such arguments include "it's invasive and creepy", but businesses seem to find enough benefit in it to spend a bunch of money on it anyway.
Unless they pull something similar to Europe, and legislate it out of existence. (Though I can't imagine how they would do this without banning security cameras altogether; you can always post-process a video feed.)
"Businesses find it useful" is a prerequisite for "people demand it to be legislated against".
Also worth noting, nobody cares about cookies any more. Nobody will care about this in 20 minutes, either.
But won't someone think of the children?
You can work out how long shoppers wait at tills by simply asking the staff - "was the shop busy today? Were many people stood in queues?"
And if the shop was busy, the answer was "Yes, and I have no fucking idea, I was too busy trying to check them out." You can't exactly plug in "kind of" and "a lot" into a spreadsheet and get useful feedback.
I'm not sure their tracking system could tell the difference between...
No, but you can adjust for that, statistically. Especially if there aren't any bus stops.
You know, employees with their own phones will really skew their stats
Unless you ask the employees to opt out. Or correlate those phones with the employee schedule. Or just edit out anyone milling around the store for eight hours at a time.
No, the problem is that this is a really crap idea that will be dead in a few years
You may be right; I think it's more likely that other technology will be used to supplement this. I bet you could already replicate most of this with a couple of kinects mounted on the walls, and a couple of HD cameras with facial recognition (as you pointed out.)
But this sort of data is really valuable to physical stores, so while this specific implementation may not last too long, in principle this is going to stay around forever, barring legal interference.
Also they could correlate purchase records with MAC addresses. That's not useful if you pay cash or only go there once, but if only one MAC address was present for all your credit card uses, they have your name and whatever else comes with that. They can locate your location in the store fairly precisely, how long you stand in front of various displays, and if they instrument the register area, they can probably get your cash purchases hooked into your record too.
You need not associate with the AP at all. If wifi is enabled, smartphones tend to do active probing for ESSIDs they want to connect to rather than turning the receiver on for long enough to listen to beacons and passively detect a suitable AP to connect to. The reason being is that it's cheaper (in terms of battery power) because they're spending less time with the receiver powered up.
Additionally,with a reasonably good deployment of APs, you can reasonably determine the location of a station (client device) down to a 3 meter level or so, this is fairly standard technology, prior to it's application to retail store analytics.
There's a similar system (known as Path Intelligence) in use in some shopping centres in the UK that tracks punters based on GSM/3G transmissions rather than connections to wifi networks, with a claimed accuracy of a couple of metres.