One-quarter of retailers are using facial recognition technology in order to track customers around their stores – with 59 per cent of fashion retailers deploying the technology.
It forms part of a widening array of potentially privacy-busting methods that retailers are resorting to in a bid to both find out more about their customers – their likes, dislikes and how they navigate stores – as well as to provide a more personalised service in-store.
On top of that, 46 per cent use facial recognition for security purposes – for trying to identify known shoplifters in a bid to target their security resources more effectively.
The new was revealed in a survey by services company CSC, in research released today entitled “Next generation in-store technology: Where do shoppers and retailers stand?”.
David Baldwin, industry general manager at CSC’s UK retail business, admitted that many customers might find some of the techniques being adopted in retail as “creepy”, but added that customer analytics would be central to conventional retailers’ efforts to compete against online – where customers are routinely tracked and profiled by retailers.
“Having the right technology and infrastructure in place, such as big data and customer analytics, will help retailers to better understand their clients. However, this data needs to be laser focused as they look to adapt to changing consumer buying habits in this rapidly evolving environment,” said Baldwin.
He added: “The survey points out that customer analytics are key to retailers providing a more personalised service to shoppers, but retailers are often overwhelmed with data.”
Indeed, in many cases, he said, retailers are scooping up vastly more data than they can use, in the expectation that it may be usable later as big data analytics develops. “They need to start thinking not just about what they are collecting, but their policies in order to make people more accepting,” said
One key area that retailers need to take more care about in terms of gaining customer confidence and consent, he added, is the use of bought-in data from third parties, and the way in which it is applied to existing databases and used – especially in an increasingly “big-data world”.
“Retailers today are collecting info that might or not be relevant but they are hoping that they can find a use case in the future,” said Ramanan Ramakrishna, director and regional general manager – UK, Ireland & Netherlands, emerging technologies.
But Ramakrishna warned retailers that they need to improve their data governance and privacy policies surrounding the use of such data in order to “make consumers more comfortable” over retailers’ increasing use of such technology.
In response to questions filed by Computing, the Information Commissioner’s Office suggested that the lawfulness of using facial recognition in-store, as the law currently stands, “would depend on how the customers personal information is being used as to whether their permission was needed”.
It continued: “If the data is used anonymously then retailers wouldn’t need permission, but if the personal data could identify someone and was being stored, a retailer would have to comply with the Data Protection Act, processing data fairly and lawfully, being transparent with customers as to how their data is used and obtaining consent where necessary.”
Facial recognition is largely used by law-enforcement agencies. Earlier this year, Police Scotland admitted that it was now using facial recognition technology in a bid to identify individuals from CCTV cameras and other sources, as well as maintaining a database of photographs as source for the practice.
The technology was also trialled in Scotland more than ten years ago by Grampian Police – one of the regional Scottish police forces that existed before they were merged to form Police Scotland – back in 2003. That pilot project, though, was aimed at searching for suspects’ pictures in a database, rather than using image recognition in the field.
The FBI in the US, meanwhile, has been building a facial recognition database of miscreants, which ought by now number some 52 million images.