Data, the new blood

Leveraging data is a vital part of any modern business strategy. Data can be used to streamline processes and gain valuable insights into customers, their beliefs and behaviours. It can even form a revenue stream in and of itself. With that in mind, it’s little wonder that many businesses hoard vast quantities of data harvested from their customers, usually as a condition of service.

Until recently, this data has remained locked away in industrial silos. Recent legal developments such as the right to data portability threaten this status quo, enabling personal data to be accessed and moved. Many researchers hope that individuals will make use of these new rights to bestow access to personal data to drive research in the public, rather than commercial, interest. But is charity a viable model when it comes to personal data and, if so, what exactly will it take to get people donating their data to worthy causes? What would it take for you to donate your data?

Leaving data solely in the hands of business and industry leaves many areas of valuable research untapped. Given the scale, variety and size of commercial data sets out there, analysing and identifying patterns across them can yield impressive and, perhaps, unexpected results, such as identifying the lifestyle causes of rare diseases and details of how diet can impact health and assessing the impacts of population-level policies. Such insights could be used to develop recommendations for health interventions or form the lifeblood of novel policies. With such a variety of possible benefits, many researchers are naturally keen to get a hold of as much commercially held data as possible.

Fortunately for researchers, avenues are starting to open up that would allow them access to data that was previously exclusively in the commercial domain. Around the world, laws governing data are starting to hand power to data subjects, including Europe’s GDPR, Australia’s Consumer Data Right, and California’s CCPA. Notably, many of these laws and regulations include a right to data portability, giving individuals the right to access and move their data. Researchers hope that they will use these new powers to allow those working in the public interest access to their data.

Anya Skatova, a researcher in the University of Bristol’s School of Psychological Science, is doing just that, and is working to make data donation an activity like any other common act of philanthropy – like giving blood. She says public health research is a good example of the kinds of public-interest work that could benefit from or become possible when researchers are able to access commercial data. “It can be really useful for public health,” she says.

Skatova says she has been focusing recently on using commercial data to shed light on public health problems. She has been working with transactional data, which includes supermarket loyalty cards and banking data. While this data could be used to illuminate any number of public health problems – Skatova says it’s especially useful in research involving diet or alcohol consumption – it’s currently used and collected by private companies to meet their business objectives, such as tracking product sales, targeting promotions and gaining better insights into their customers.

Skatova says donated commercial data, such as that from loyalty cards, can overcome some of the shortcomings of other research methods. It offers a new way of studying behaviours that can otherwise be notoriously difficult to study. There are all sorts of biases, over-reporting, and under-reporting with behaviours like alcohol consumption and diet that can really hinder researchers’ ability to understand what’s going on, Skatova explains. “These types of behaviours are usually studied by self-report,” she says. “They’re really difficult to study because of these biases.”

Skatova is keen to stress that donated data will not replace existing research, but will add a new dimension to work and provide another way of gaining insight into behaviour. “It can provide a different insight . . . but it won’t replace [existing research methods],” she says.

The benefits to public health and other areas of research evident today are only the beginning of what could be possible. It’s still early days on seeing what benefits or possibilities there might be when using donated data, Skatova says. “It’s ongoing work with data holders and trying to understand just how this data can be useful for public health,” she says. “There isn’t a route yet for somebody to just go and donate their data. We’re working on it, though.”

Laws empowering data subjects to access and move their data are only part of the picture when it comes to data donation. Absolutely vital to getting any data donation effort off the ground is persuading data subjects to exercise their rights over their data and actually donate it.

Donating data is similar to other forms of altruism, such as volunteering time, donating blood and charitable giving. “We see analogies to blood donation or any other charitable behaviour,” says Skatova. “This is a new example of prosocial behaviour people might engage in more in the future.”

The good news for researchers hoping to unlock the value in commercial data is that Skatova’s research shows that many people are actually quite willing to give away their data for good causes. In a recent paper published in the journal PLOS ONE, Skatova, along with colleague James Goulding, found that over half of individuals were willing to donate their personal data for research that would benefit the wider public. Reasons for donating, they found, aligned well with those already established for other acts of philanthropy, such as giving blood.

Unfortunately, even though the research indicates that many are willing to donate data, there are a number of barriers that stand in the way of data donation becoming a widespread practice.

Perception

Knowing that people are willing to donate their data to researchers doing work for the benefit of the public is, unfortunately, only a fraction of the way to getting data donation programmes off the ground. All other barriers and obstacles aside, Skatova says that the most significant hurdle comes down to actually persuading people to part with their data, and convincing them that a particular project is one for the public good, rather than one where their data will be misused or exploited.

The uses and abuses of personal data for political targeting are perhaps the activities that loom largest in the public eye when it comes to personal data use, with the Cambridge Analytica scandal being the most prominent example. This, along with a seemingly endless slew of high-profile, very public cases of unethical or problematic uses of personal data, has understandably fostered a culture of severe distrust and suspicion towards any group or entity seeking to make use of personal data. To many, the prospect of exercising their recently acquired legal powers over personal data to give data back to someone else is almost laughable.

“There’s a lot of danger in the public eye,” says Skatova. “These risks are really perceived, though. They’re not really real.” She says it can make working with individuals and companies alike an uphill struggle. “It’s been really difficult to talk to supermarkets and other data holders [about using the data] because of this kind of public perception factor.”

“It is so deeply ingrained in peoples’ minds that this data is used by companies for profit. So when we do focus groups we always need to argue and defend ourselves that we’re not doing it for profit even though when we come from a university. Of course, some people do research for profit, but it’s a minority.” Talking about loyalty card data, for instance, often has people thinking about marketing or other for-profit uses, Skatova says. “It raises all these red flags.”

Skatova says that given the public perception over how personal data is used, it is important for herself and other researchers to be crystal clear that their research is for the public good. She hopes this public perception will change, with a lot of beneficial research being hampered, or entirely unexplored, through lack of data. “What I really want to do is create some kind of public narrative that this data can be used for public good without it being used for profit.”

“I think the fear is that we will develop something that will be published and accessible by anybody and then that will be used by somebody for profit or something bad,” she says. “But then it’s the same with all research. It is for data research, or medicine, or nuclear physics. So there is a tension there and I really hope we overcome it. Because of these tensions we have all this data locked in silos in companies and they’re doing research for profit, yet we can’t get it for the public benefit. That’s why data donation is really important.”

There are a lot of legal grey areas left surrounding data donation. The issues are not unique to donation and can be deceptively complex. The law around data ownership is one such grey area, with a great deal left to work out in various jurisdictions determining who can own a particular dataset, under what circumstances and what happens to products or research derived from that data.

Another particularly important grey area, says Skatova pertains to the responsibility companies may or may not have over data that subjects have asked to be transferred. In some instances, she says, it’s possible companies may still have various obligations to keep data safe even if subjects ask to release it.

Another problem is infrastructural. A great deal of commercial data that subjects can now request was not designed with sharing in mind. This can cause problems for researchers seeking to make use of it further down the line, even if they have been able to gain access. Skatova says merging data sets, such as those from transactional or loyalty card information, from different sources can be challenging, making research more complicated.

The actual mechanisms of requesting and sharing data can also prove problematic. There are presently no standards here; even where legislation provides for data portability, many organisations are still themselves trying to establish processes to share data. With no standard, simple way of acquiring and transmitting data, the burden can fall on subjects, who may not go through all the extra effort to donate.

There are few incentives for companies to take the lead and be the first to set up data-sharing standards, Skatova adds, which further complicates matters. Companies are wary, she says, of becoming the next poster child for personal data misuse should something go wrong.

image
SHUTTERSTOCK.COM/ ABSCENT, HAPPYMAY

A different barrier comes from the academic nature of the research. Unlike a great deal of business or commercially run research where data has been acquired in the course of business, academics often have to deal with stringent codes of ethics. A lot of research in the public interest, such as public health or behavioural research, often finds itself under the aegis of some of the stricter medical ethics frameworks. Such codes throw up hurdles for good reasons, but not all are appropriate for work with big data. Informed consent, the bedrock of medical and behavioural research, is not always possible with data-driven work where outcomes may be uncertain and hard to anticipate beforehand.

Without resolving many of these issues, getting data donation projects off the ground remain uphill battles. “A lot of these issues sound boring, but they’re the practicalities we’re dealing with,” Skatova says.

Looking ahead

As it stands today, researchers’ hopes for data donation research programmes remain just that: aspirational. As in the business world, meaningful data-driven research relies on masses and masses of data. Unlike in the business world, public interest research is not able to rely on the inertia of data subjects; it requires active engagement, assent and participation to get projects off the ground. This means that now the legal instruments are in place to enable data subjects to share personal data, the challenges remaining for researchers are persuading subjects to do so and establishing the means of doing so easily.

Skatova feels that tackling some of the public fears surrounding data will go a long way to achieving those goals. “It’s just changing public opinion,” she says, adding that a greater public awareness on the beneficial, non-exploitative uses of data could encourage companies to be less fearful in developing data-sharing infrastructure.

As with other kinds of prosocial behaviours, Skatova questions the roles charities might play in data donation moving forward. “I always think that an ideal situation could be if there’s a data charity that you can donate your data to for research.” Another option, Skatova says, could be for existing charities to act as data banks for particular types of data. “At the moment they’re not really involved in this,” she says. “But they should be. This data could be really useful for them.”

Get unlimited access to all Global Data Review content