A British company Pixoneye has developed technology that scans a smartphone’s photo galleries and generates consumer insights based on what it finds.
The software behind Pixoneye integrates into a host app, like your members app for instance, and asks the user’s permission to process their photos. Using a process the company calls ‘image understanding’, the algorithm then ‘reads’ the photos. It looks for anything useful, such as the user’s age and interests to whether they have pets or children.
That info can then be used to push highly targeted ads and special offers through the integrated app.
There’s a lot of information in personal photographs. Pixoneye Chief Executive Ofri Ben-Porat says their software can retrieve over 150 unique characteristics from users’ photos. The basics include age, location, gender. The more surprising claims include the ability to determine income level, fashion sense and holiday style.
“Smartphone photos are the closest we have to offline, personal data,” Ben-Porat said. “Out of a gallery of 1,200 photos, the user maybe uploads 1% to social media to keep the façade alive. But it’s about what happens to the other 20 photos you’ve taken, or shared internally, which you don’t look at again.”
Access to that level of data means marketers can fine-tune their targeting and needn’t rely so heavily on carefully constructed social media profiles.
Any photo-scanning app will come with privacy concerns. Porat believes Pixoneye’s software respects user privacy completely.
“Pixoneye does not capture or store any images, and instead interprets objects as mathematical data,” said Porat. “It completely disregards any object or face. It looks at angles, textures, colours and breaks those down into a formula.”
According to the start-up, all information processing happens on the device. This means Pixoneye doesn’t have to send any photos to—nor store any photos on—its own servers.
Currently, Pixoneye’s capabilities must be integrated into a host app. They offer a software development kit (SDK) that they say will work with any platform.
It’s a very interesting concept from a club or pub point of view. We could certainly find out more about members interests and make far more highly targeted offers and personalised communications. Watch this space – this is the next level up from location-based marketing and I’m sure we’ll see applications appearing on the local market soon.