TORONTO: Technologies such as artificial intelligence threaten to make many traditional forms of market research redundant, according to a leading executive from Sqreem, an analytics firm.

Ian Chapman-Banks, CEO of Sqreem, discussed this subject at 2017 Global Marketer Week, an event in Toronto co-sponsored by the World Federation of Advertisers (WFA) and the Association of Canadian Advertisers (ACA).

"I feel sorry for Ipsos and Kantar and all these guys," he said. (For more details, read WARC's free-to-access report: Artificial intelligence and the future of consumer insights.)

Specifically, he stated that Sqreem's data-acquisition engine can now "take 85% of everybody's digital daily activity", from search to web browsing, providing more in-depth and rapid results than established research techniques often muster.

"Do you really want to go and interview 40 people to understand and model up to 290 million people in America? … It takes them four months and $250,000," said Chapman-Banks.

"Years ago, you had chimney sweeps and secretarial pools and people in elevators pressing buttons for you. That's the market-research agencies."

Sqreem, he continued, is able to analyze consumer habits at huge scale. "In about 40 countries, we behaviorally profile everybody," Chapman-Banks said. "We've actually discovered there are 450,000 things that bring us all together.

"It can be 'I need to go shopping', 'I'm too fat', 'I'm depressed', 'My wife hates me', 'I need to get promoted', 'I need to buy life insurance', 'What's my next car going to be?'"

With that information in hand, it is possible to target people based on their observable behavior, Chapman-Banks told the WFA/ACA assembly.

"If we take those 450,000 behavioral profiles, we can get down to audiences – if we wanted to, which really we don't want to – of hundreds of people, and we can target those people," he said.

"We have an automated engine where we actually take those insights and put them directly onto Facebook, or LinkedIn, or whatever you want to do."

Data sourced from WARC