Media investment firm GroupM has launched a first-of-its-kind tool that attempts to measure the ethicality of decisions in the advertising industry, using dozens of individually weighted factors to determine whether a move should get the all-clear or warrants some revisions.

Created in partnership with Unilever to address the growing disparity between “good data use” and “bad data use,” the new Data Ethics Compass lets advertisers quickly evaluate the ethical risk level of data-based decisions using “proprietary scoring logic” that the WPP-owned group says removes subjectivity. 

(That’s half-true. As the Compass was crafted from WPP’s ethical point of view, another agency or brand might arrive at different moral conclusions based on their in-house beliefs, and even the tool’s creator, Krystal Olivieri, concedes that every ethical decision must be processed and weighed individually to determine risk.)

For its initial roll-out, GroupM has tapped its media agency Mindshare to apply the Data Ethics Compass to some of Unilever’s data-driven media activations.  To refine it, numerous active brand and campaign scenarios have been analyzed to translate the tool’s ethical frameworks into regular, widespread use.

Here’s how it works

Broken down into five primary stages, the Data Ethics Compass first asks users for a wealth of contextual information, including their organization’s name, their location, their office’s location, the location of the consumers they intend to target, and more. 

The reason for this in-depth background is because “when people are flying through these tools, we don’t want any ambiguity,” explains Olivieri, who is also GroupM’s global senior VP for data strategy and partnerships.

Users can then choose their business sector; retailers, charities, gambling corporations, political parties, TV broadcasters and a host of others are represented. “We start with the client sector to contextualize, because different client sectors have different relationships with end consumers,” Olivieri says, giving the example of a travel company having more information about a consumer—and using it differently—than a health care provider.

Step three involves identifying the source of a user’s input data; is it first-party or third-party? Choose one, and then specify further with a drop-down menu of choices. After clicking on third-party data in this step, for example, clients may pick from source options such as audience data, social platform data and panel or survey data.

After narrowing the input about how the information was collected, users can select a “data type category”—age, marital status, criminal history, sexual orientation are all in play—to reveal what a specific factor says about a consumer. Again, this may prompt a drop-down menu; if you select location, you can then add the type of data it has been gathered from: GPS coordinates, ZIP codes, IP addresses and so on.

In future versions of the tool, clients will be able to pick multiple options if needed to specify intersectionality, Olivieri says, but in the initial roll-out, only one data type category per analyzed decision is supported.

Then, users are brought to the fifth and final step, where they can pick how their data will be used. Are you aiming to work on customer behavioral targeting, or is gaining demographic insights your objective?

After diligently following those steps, users will receive an ethicality rating for their decision on a dial scale. An arrow in the green means WPP, in accordance with its own moral perspective, sees little to no ethical risk; yellow indicates an “acceptable” level of risk; orange means more information is required; and red raises the alarm, prompting users to rethink their high-risk scenario.

What does it mean for marketers?

“This is a human decision-making tool. So, we’re not running AI or machine learning to build this, we’re actually sitting together and working through this with our data science teams,” Olivieri stresses, noting that the tool’s ethics composite is open to interpretation on a case-by-case basis.

An orange or red rating on the Data Ethics Compass isn’t necessarily grounds for WPP to dismiss a decision, either. If a result is on the riskier end of the spectrum, clients can cite the report card that accompanies their ethical risk score and revise their strategy appropriately. 

“We use the terms ‘ethical risk’ because we’re not saying a tactic is ever ethical or not ethical. Ethics is subjective,” she says. “But we are saying that we see some level of risk from a consumer’s perspective within [any given] tactic.” And increasingly, consumers are staking great interest in the ethical positions of the brands they support.

Some factors are more likely to trigger a red-level risk warning than others, she says, giving the example of data “fingerprinting”—a tactic that WPP's playbook has ruled a no-go. But at the end of the day, it’s up to the user to decide when and how to implement their decision.

“The burgeoning of privacy protection laws around the world reflects the widespread consumer demand to have the interests of the individual paramount in considerations concerning data usage,” GroupM general counsel Nicola McCormick says.

For consumers, ethical considerations are up to three times more importance for fostering corporate trust than competence, meaning simple black-and-white decisions about “opted-in” data will no longer suffice in an industry with increasingly complex data privacy practices. 

The Data Ethics Compass enables GroupM and the tool’s users to “think more holistically” about their data-driven decisions’ impacts on their targets, she adds.