February 28, 2020
February 28, 2020
Contributor: Laura Starita
Organizations must define their ethical framework for proper data use before they pursue new data-driven initiatives, and regularly after launching them.
Should a K-12 education system collect data from school laptops to assess student performance? Or a car manufacturer collect and share data on vehicle trips to inform smart city programs? Or should a maker of sensor-equipped industrial equipment use IoT data to make product improvements and sell insights?
Organizations could argue that collecting these types of data enables them to improve service and offer broader social benefits. Conversely, students or drivers or customers could argue that using this data violates their expectations about how data is collected and used.
Data ethics dilemmas like these are becoming more urgent as business leaders look to data and analytics programs to produce business value.
“As more organizations look to benefit from data, there will be an inevitable increase in data use and sharing missteps,” says Lydia Clougherty Jones, Senior Director Analyst, Gartner. “Organizations with an ethics culture will be better prepared to avoid missteps altogether or handle them effectively when they occur.”
Gartner defines “data ethics” as a system of values and moral principles related to the responsible collection, use and sharing of data. Data ethics violations range from overt and public to subtle and secret — like algorithms that suggest higher interest rates for minority mortgage applicants or lower lines of credit for women credit card applicants.
Whether overt or subtle, ethics missteps are bad for business. Affected stakeholders are left feeling that a promise — implicit or explicit — has been broken. Explanations based on after-the-fact analysis appear, when viewed through an ethics lens, disingenuous — even hypocritical. For organizations, the crisis is often a surprise, as few consider ethics until there’s a problem. Only then do they stop to consider the potential risk of unintended consequences.
There’s another way. For a more proactive stance, take these three actions.
Many organizations approach data governance as a set of hard-and-fast rules, guided by data protection regulatory requirements like the European Union's General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA). Unfortunately, a fixed approach to data governance is ill-suited to today’s rapidly changing digital environments. Data ethics also encompasses far more than privacy and security compliance.
Shift the governance mindset away from command-and-control or one-size-fits-all toward adaptive governance instead. With adaptive governance, organizations determine the right governance styles and mechanisms for a given context.
“Data-sharing strategy for intentional impact must exist within an adaptive ethical framework to balance risk with contextual opportunity,” says Clougherty Jones. “Building trust with emerging technologies is essential, along with aligning data and analytics strategy to organization goals.”
An adaptive approach invites leaders from different functions to collaborate on data ethics principles to guide decision making, while acknowledging that new circumstances may arise for which there is no precedent. Thoughtful leadership discussions about what it means for the organization to do the “right thing” helps provide a framework for ethical thinking and decision making.
Organizations need an active and vocal data ethics leader at the helm with authority to drive ethical standards. Although many organizations have multiple people handling ethics, data and analytics leaders can establish themselves as data ethics experts.
One key task will be to encourage cross-functional conversations about data and appropriate use and sharing with peers in marketing, technology, product development, finance, legal and across the business.
Start these conversations early in the process when planning new data initiatives or projects, and revisit the discussion at frequent intervals to address ongoing or new challenges. Focus on the ways that an ethical approach can cultivate customer trust and drive business performance.
Spend time discussing and grappling with complex notions of right and wrong to develop a cultural view of data ethics. Given the contextual nature of data and its use, employ real or realistic dilemmas as jumping-off points to fuel discussion. Also be mindful of customer and stakeholder norms and expectations around data use and sharing.
Consider the example of a retailer that wants to capture customer location data to share with business partners and to send coupons when shoppers near a store. Three mental models could drive how the organization thinks about ethics:
Each mental model can be used to justify a yes or no position. A universalist might look at company policy and decide that it’s ethical to send a coupon if the company allows location data and it’s unethical if it doesn’t.
A consequentialist may say that if the coupon benefits the customer, then sending it is ethical, whereas it’s not if customers find unsolicited coupons creepy. Care ethicists might say that opt-ins or other permissions make unsolicited coupons okay or they may conclude that there are other, less-intrusive ways to attract people to stores.
Discussions to unpack the complex nature of real-world data decisions help prepare organizations to do the right thing all the time — not just when something bad happens.
Join your peers for the unveiling of the latest insights at Gartner conferences.
Recommended resources for Gartner clients*:
Data Ethics Enables Business Value by Lydia Clougherty Jones, et al.
*Note that some documents may not be available to all Gartner clients.