Posting in Technology
A new startup considers itself 'Google analytics for the real world' by tracking customer browsing habits through security cameras in retail.
Security cameras in retail outlets are an accepted part of the shopping experience. You expect to be monitored, and the footage is generally only used to bring shoplifters or those that cause a public disturbance to justice.
However, the vast amount of shopper data which is available due to these cameras may also soon have other purposes. In the same manner that online stores, such as Amazon, track customer preferences, habits, and alter their business strategy or recommended products based on this data, perhaps brick-and-mortar stores will now begin to follow suit.
Prism Skylabs, based in San Francisco, now allows retail owners to mine extra information from their shop footage through the use of new scanning software. The company installs the software on computers that link to a store's security camera system, compresses the gained information, which is then sent to cloud servers. Once the raw data is uploaded in to the cloud, Prism sends back extracted data, including statistics and visualizations.
This can be used to count, log, and track people in-store. The software is able to analyze customer density, where shoppers deviate most to when they are present in the store, and track browsing habits. Anything from the length of a checkout queue to recording the most popular products or showrooms can be tracked and analyzed in retail through normal CCTV footage.
Steve Russell, CEO of Prism says:
"There's a lot of wonderful information locked up in video, and 40 million security cameras in the U.S. collecting it, but it's data that's not been available. We want to free up that information."
Although the software turns the usual low-resolution footage available in to a higher resolution version, in order to maintain personal privacy individual features are blurred. Human figures are given a 'ghostly' image, or removed completely, leaving colored spheres in their place. This can also be used to generate 'heat maps' to track crowd density in specific areas, and give retailers more of a glimpse in to the patterns of their customers.
Image credit: Shaun Greiner
Related on SmartPlanet:
- Swap-o-Matic: Consumerism to cast-offs
- Tesco abandons carbon-label initiative
- Can mobile wallet tech save retail stores?
- Web-savvy consumers spend more in stores (research)
Feb 2, 2012
Security cameras in shopping retail store are very important part. It help in many form for different purpose.
I really doubt that "tracking" what I buy on any given day can predict what I might buy in the future.
interesting and yet disturbing these applications of the cameras invade privacy in a very subtle. Security Cameras
What is to prevent unscrupulous store owners and/or software authors from analyzing camera data gleaned from cameras aimed (for instance) at men's and ladies room doors. They can then determine if a particular shopper is spending too much time in the john and target them for laxatives, diuretics, toilet paper, etc How about cameras aimed at the sanitary napkin aisle, or the condom shelves? That might make a few shoppers a bit queasy!! What we need is a clever hardware engineer to come up with a camera-jamming device, just like the one that now can be used to remotely change a TV channel, turn down the audio, or turn it off!
Good plan. Process could crowd source store layout optimization. Privacy is just something we will have to get over.
I have Hulu-Plus which ALWAYS wants to improve my "Ad Experience" and Netflix, which is trying to prognosticate my tastes based on what I just watched. I don't make buying choices based on Commercials and my tastes are WAY too eclectic to be analyzed. Until I simply got tired of even bothering to interact with Hulu's "Ad optimization" App, I would change my answers and options on a regular basis. About the only thing I do when shopping is "Price & Support" compare. Other than that, any attempt to confirm a pattern is not only pointless, but likely to skew the outcome. I suggest this is likely happening with ALL this data, as humans are know to be a rather capricious species. As this skewed Data adds up, the forecast models will become evermore useless. As I think of the folks who will be buying into Facebook's $100M IPO, I can't help but snicker.
When you create another use for the Video Content (beyond Security) now you have a real ROI. One that makes sense. Tracking is nothing new - some call it analytics. But only until recently have these tools been available to a more wider audience.
I want to do a consumer study on how retailers collect information and want to station cameras outside the stores. Creepy or what! I have no problem with security, I don't want an unknown company taking unknown information and doing unknown things with it.