October 27th, 2016
‘We just need to train AI along the way’: xAd CEO on the future of location advertising
Location-based data is growing in sophistication, thanks to advances in Big Data and machine-learning. It also presents marketers with new opportunities to reach and target consumers in real-time, building deeper and more accurate social profiles.
One leader in location technology is xAd. In his presentation at the company’s inaugural Emerge Mobile event in London, xAd’s chief executive officer Dipanshu Sharma said: “Location is the most accurate indicator in identifying an audience. And there are enough devices in our pockets that define us better than a social graph.”
But how many brands are taking advantage of the information being collected by these mobile devices? ABI Research found that retailers are failing to exploit the opportunities of mobile and indoor location technologies and warns that retailers will lose control of the mobile experience in their stores if they don’t act now.
xAd enables brands to tap into the mobile space by leveraging location-based data to target consumers with ads in real-time. Sharma started the location-based marketing company seven years ago and since then it has grown to hundreds of employees. It’s also the only location company to be featured alongside companies worth over $100m in revenue in LUMAscape, the industry standard for understanding the digital media economy.
Speaking to The Drum after the event, Sharma says the easiest way to explain what xAd does is to compare it to how Google indexes websites – except xAd indexes the “physical world” to learn more about people’s behaviour when they are offline. And he says xAd’s relationship with brands is mutually beneficial.
“We’ve got location data to do with what is happening, they’ve got data on what people are buying. If you bring them together, then it’s so much more powerful than keeping them separate. A lot of brands are having that dialogue with us,” Sharma explains.
Sharma says xAd is collecting data on 500 million consumers every month, billions of times a month which he said is central to the machine learning that underpins xAd: “Data is the new currency, the new oil, like a lot of people are saying today,” he says.
The problem with earlier versions of location-based technologies, says Sharma, lied in their inability to tell whether a person was near a building or actually inside the building. A very important difference, as it contributed to building inaccurate user profiles.
“If you think about somebody that is at a dealership looking at a car versus somebody that is driving by the dealership, that’s a very different context,” Sharma explains. “We saw a lot of noise in the signal and that created inaccurate profiles of behaviour patterns.”
Sharma says the new version of location technologies are polygon based, like xAd’s Blueprint technology that solves this boundary problem: “There’s a sporting goods store in the US that caters to hiking and camping and so forth. We started creating boundaries for camping grounds to see who really goes camping, and we captured their profiles and started targeting them with ads for sporting goods stores. The more polygons we create, the more accurate our audience profiles become.”
But as AI becomes more sophisticated, there is a danger of getting too personal. Most users don’t understand or have the time to figure out how their data is being collected or used. But xAd does not focus on specific individual data and has several safeguards that protects the company from identifying individuals. Still, Sharma agrees that there is a certain responsibility that goes with “choosing the creative sensibly” and admits that brands will not always get it right, making “mistakes” along the way. He thinks it’s important for humans to always be involved in the process and to constantly probe and ask the right questions.
“Eventually AI will get better, we just need to train it along the way,” Sharma laughs.
Is there anything Sharma is particularly excited to see in the industry in the upcoming years?
“What Pokémon Go did was bring location and augmented reality together which was really exciting. The future is predicting where people will go next. Today we know where you are and where you were. And then we use that data to do a good job at targeting you with the right ad.
“In the future we will know where you are going to go and you will see the ad before you even make the decision, because we will have so much data on what you are doing on a daily basis. I think we will start to see this from 2018 onwards, and probably start seeing some really good technologies around that,” Sharma concludes.