#RetailEXPOVC: Use data but always remain sceptical, says Dr Hannah Fry

Retailers and the retail community were treated to a deep dive into data during the opening keynote session on day two (30 April) of the RetailEXPO virtual conference.

Dr Hannah Fry, associate professor at University College London, and celebrity mathematician on programme's such as BBC 6Music’s Breakfast Show feature 'The Maths of Life', urged retailers to use data but remain sceptical of it.

In an illuminating session that gave examples of how the work of Dr Fry and her colleagues is aiding police investigations, and helping shape some of the potential Covid-19 post-lockdown government strategy, she offered plenty of relevant advice to the retail and business community.

“You have to be sceptical all the time and not get too excited about interesting results,” Dr Fry explained, in reference to analysing data that many retailers use every day to understand customers, plan business processes, or map out future strategy.

“I think it’s about embracing uncertainty and remaining sceptical.”

More content from the RetailEXPO Virtual Conference, powered by Essential Retail:

#RetailEXPOVC: Five ways retailers can adapt to the Covid-19 lockdown

#RetailEXPOVC: Vivobarefoot on regenerating retail to save the planet

#RetailEXPOVC: Ex-Waitrose boss on the post-Covid retail world

#RetailEXPOVC: Five ways retailers can adapt to the Covid-19 lockdown

#RetailEXPOVC: How Pizza Hut regained its appetite for digital

#RetailEXPOVC: Five mobile strategy tips from The Very Group

#RetailEXPOVC: A fresh approach to marketing during Covid-19

#RetailEXPOVC: Three ways retailers can encourage women to smash glass ceilings

Answering questions directly relating to GDPR – the regulations now governing how retailers and other organisations should be handling customer data – she added: “I think we’re still trying to find our way, people haven’t really settled in on what the real interpretations should be.”

“The thing that’s promising for me in terms of GDPR is we’ve seen a big shift in people coming up with techniques that preserve people’s privacy without preventing innovation. [There are] lot of things like differential privacy – I think that’s the stuff that’s been really positive.”

Differential privacy makes it feasible for technology companies to collect and share aggregate information about users or consumers, while preserving the privacy of individuals. It is used by the likes of Apple and Google in their respective operating systems, and has an array of potential areas of use in retail, from payments to marketing.

Machine learning and top tips

Reflecting on a year she spent working closely with artificial intelligence (AI) company, DeepMind, Dr Fry said AI companies would argue there is nothing about human brains that’s so special it cannot be replicated by machine. If you can teach someone how to do something, then you can teach a machine how to do it, is the theory.

“Personally, while that might be true, I think we’re a very long way away from it – far enough away from it that it might not be true,” Fry noted, advising caution when using machine learning techniques in business.

Indeed, she added that many algorithms need a guiding human hand all the way through, as no matter how good they are “they don’t really understand context or nuance”.

Other titbits from the session include Fry’s advice that “you don’t necessarily need vast amounts of data to tell you something about human behaviour”, and that when analysing patterns in human behaviour “you have to take into account the weirdness of humans”. Essentially, there will always be anomalies.

Sage advice for retailers and brands, it seems, as data analysis and science become more common in their everyday operations.