Putting Big Data on the menu
The screens in our pockets have become a commonplace tool for the research, selection and purchase of everything from supermarket groceries to fully cooked meals, delivered straight to our doors. It has some health experts worried, but others see an opportunity to use the wealth of data generated online to create better health outcomes.
Household budgets, supermarket design, time pressures, product packaging, nutritional guidelines, marketing and advertising – all of these factors come into play when we make decisions about the food we buy, the food we eat and the food we share with our families.
The connection between socio-economic status, food choices and health is well documented, and Australia’s rising economic inequality will certainly impact on the health of the population in the years ahead.
But the effect that digitisation will ultimately have on the health and wellbeing of the population is also being evaluated by public health experts like Dr Sandro Demaio, fomer Chief Executive Officer of EAT, a science-based, global platform for food systems transformation.
‘Digitisation has been fundamental to improving the lives of millions of people worldwide,’ he says. ‘It can soften the equity gaps in terms of access to information, job markets and healthcare. It allows us to better track health treatments and ensure people are getting consistent healthcare.
‘But it also comes with some compromises and challenges, and one area that is very important, particularly from a public health perspective, is the digitisation of our food supply system.’
Consumers choose convenience
Apps that support speedy home-delivery from a wide range of restaurants and fast food outlets are among the most novel, and yet most familiar elements of this new digital food ecosystem. The sharp rise of the ‘aggregators’, including the likes of Deliveroo, Foodora and Uber Eats, has dramatically changed the restaurant and fast food markets. A report from Morgan Stanley estimates that online food delivery platforms in Australia are generating around $600 million in total transaction value and anticipates this increasing to $2.4 billion by 2026.
‘The attractiveness of home delivery is clear,’ says Demaio. ‘Consumers can order food from the convenience of their home and peruse an expansive array of food options.’
Traditional supermarkets are looking for their share of this emerging market, too, delivering ready-to-eat meals directly to our doors. In the US, Amazon Go is offering hyper-convenience (if not home-delivery) via self-service, virtual-checkout supermarkets that prioritise ready-to-eat meals.
Demaio makes the observation that convenience is driving food choices like never before, but the shift from home-cooked meals to ‘instant meals’ may have nutritional consequences. The issue is that what presents as a generous range of food choices is in fact a set of options dominated by products that deliver profitability to the food sector, often heavily weighted towards sugar, salt and fat. And hyper-personalised marketing, driven by data-fed artificial intelligence, make those compromised food options even harder to circumvent.
Highly tailored, unhealthy marketing messages
The digital food ecosystem has the capacity to leverage large amounts of data generated by legions of online consumers. Big Data allows for a sharply focused understanding of the food consumption habits of all users, starting with their demographics and geographic location and drilling down to the dietary preferences of the individual.
That highly specific knowledge can be used to commercial advantage, as profit-driven businesses tailor the individual consumer’s experience of the digital food ecosystem in compelling ways.
‘This tailoring, combined with the traditional challenges of high-fat, high-salt, high-sugar, highly processed foods will make it even more sticky for the individual,’ says Demaio. ‘My concern is that the result will be an acceleration of many of the health challenges that we face, including child and adolescent overweight and obesity, adult obesity and chronic disease.’
Demaio’s concern is heightened by the current lack of a regulatory framework for digital marketing, an issue not unique to Australia. Children in particular are at risk of being exposed to these highly persuasive and intensely targeted marketing messages.
‘It seems that policymakers are yet to appreciate the full picture of how this sort of digital ecosystem, combined with a risk factor like food, could play out in terms of public health,’ says Demaio.
"My concern is that the result will be an acceleration of many of the health challenges that we face."
Health messengers struggle to be heard
One of the impacts of global digital disruption has been the decentralisation of information. In some respects, the change has been positive and transformative, giving smaller, more marginalised groups the opportunity to have their voices heard. But traditional information sources have suffered some collateral damage, losing influence in favour of alternative sources.
This has been a challenge for public health, where evidence-based recommendations around issues such as healthy eating may be even less prominent than those of the digital media ‘influencers’.
‘It’s a much more crowded landscape in terms of the so-called experts and what that term means today,’ says Demaio. ‘We’re seeing much more noise and misinformation around what a healthy diet is, around doctors’ recommendations and just generally about food.’
In the context of ever-increasing social media use by Australians (91 per cent have a Facebook profile, 39 per cent have an account on Instagram), Demaio makes the observation that much of the health information we receive – actively or passively – comes through digital platforms.
‘Public health organisations are lagging behind the big corporates in the degree, effectiveness and creativeness with which they are engaging on novel platforms,’ he says. ‘[They] must increase and improve their presence on these platforms if they want to remain relevant and effective.’
Big data for health researchers
The question for the public health community then is how to take the looming challenges and transform them into opportunities.
‘How can we turn some of these predatory technologies to benefit public health?’ asks Demaio. ‘How can we work with online retailers to ensure that Australians are able to access better quality, cheaper, fresh food more quickly, more efficiently, and more affordably?’
"How can we turn some of these predatory technologies to benefit public health?"
Demaio likes to imagine an online interface, produced collaboratively by health promotion agencies and commercial stakeholders, that uses data-based insights to encourage healthier food choices, supported by easier access to nutritional information.
Another scenario would be for public health agencies to tap into the data already being collected by online food providers.
‘These apps have incredible amounts of data for so much of Victoria, telling us what the food environment looks like. It can tell us what the food options are for an individual in a particular street at a certain point in time.’
That data could be used to better understand the context of food choices, to identify targets for funding and interventions, and to help evaluate their impact on disease and obesity outcomes in adults and children. Or it could help to identify neighbourhoods where healthy food choices are scarce, allowing public and private operators to work together to make more healthy options available.
Demaio muses on the possibilities of even richer collaborations. Imagine, for example, if food delivery data could be used to identify restaurants and fast food outlets that typically produced less healthy meals, and to connect them with food and nutrition experts offering training and upskilling.
‘Can we think of different models through public-private partnerships or through social enterprise partnerships to bring more healthy food options to consumers?’
New digital teams monitoring public health
The broad scope of collaborations and interventions made possible by data is best met by a more expansive definition of what it means to be a public health professional.
We need to be training people from different fields, not just those with biomedical roots,’ says Demaio.
Sourcing experts and practitioners from fields such as urban planning or an industry tied directly to the digital world will increase the range and quality of data analysis available in the public health sector, argues Demaio.
He advocates for more digital training for existing public health professionals so they can better understand and evaluate commercially led online environments, and so they are equipped to conceptualise interventions that are not only effective in that context, but which take best advantage of new and emerging digital possibilities. He also puts the case for more investment in cross-sector partnerships.
We need to be humble and position ourselves to be a great partner with industry. If we can partner with, for example, digital advertising companies who want to use their skills, tools and reach to do good and to improve the health of populations, there’s a very exciting partnership opportunity.’
These new relationships may be outside the norm for public health organisations but, says Demaio, ‘the access to data, the access to food producers and providers, the access to consumers’ could provide a basis from which to positively influence the aspects of the digital food ecosystem that pose a risk.
Demaio sees the digital food ecosystem as not only the new marketplace for consumers; it’s also an important new workplace that needs to be a priority area for public health professionals.