Google to predict what consumers will buy and where by using cross-device measurement
Google has upped the stakes by linking YouTube views and other searches directly to retail store visits, in a move it claims will enable it to accurately predict when people will walk into a store to make a purchase.
The development aims to showcase the link between online advertising and in-store purchases.
Google has also announced store measurement will be linked to YouTube to measure store visits on the back of people having seen video.
The mapping technology has been used since 2014 to measure store visits resulting from internet activity, but will be extended to video for the first time in Australia.
It will also be linked to store sales measurement allowing retailers to gauge and predict exactly what customers will buy.
“The digital and physical worlds are merging,” Jerry Dischler, Google’s vice president for product management at AdWord said.
“We have shown that online advertising is driving consumers into stores. We have measured over five billion store visits globally in 17 countries that were the result of ad clicks. So we measure from a click on a search ad or a display ad all the way through a store visit.”
The system not only measures what people are looking for, but can also show them store inventory.
It is linked closely to Google Maps, but also monitors WiFi networks and other systems for activity to build a clearer picture of consumer behaviour.
“On top of that we also have five million human raters where if we are uncertain whether people are in a store location we ask them and we feed that into a deep neural network and use that in order to get 99% precision knowing someone is in a store,” Dischler said.
Google said the new systems will continue to protect privacy, but will give retailers a better picture of the pre-sale behaviour of consumers.
I assumed they were already doing this.
User ID not verified.
“Google said the new systems will continue to protect privacy”
Hilarious.
User ID not verified.