Data Ethics: 6 Steps for Ethically Sound People Analytics

Dirk Petersen of Insight222 shares his sage advice on data ethics and how people analytics teams can successfully build up this capability.

In late July 2017, a Who’s Who of Silicon Valley people analytics leaders came together at Intel. No one foresaw that Cambridge Analytica would soon raise the profile of data ethics to a new level or urgency. In fact, the leaders that day confirmed what similar groups had found at meetings earlier that month in London (hosted by Cisco), Amsterdam (ABN AMRO), and New York (Pfizer): ethical data practices are important, but are not urgent.

At that Intel meeting, the people analytics leaders from 55 companies came to the realization that people data ethics is critical because of its foundational importance for people analytics success. No organization can build a sustainable, successful people analytics function without clear understanding of what it should and shouldn’t do. Participants realized it is the foundation on which all analytics work builds.

An Insight222 survey with 57 companies* shows the extent of the challenge: 81% of companies report that their workforce analytics projects were sometimes or often jeopardized by data ethics/privacy concerns.

Even the most sophisticated companies face challenges in this area: project delays because of uncertainty whether the data gathered complies with GDPR; projects cancelled because works councils are not on board; analyses weakened because of lack of alignment among company leadership on whether ‘it’s ok’ to run certain algorithms. Those who’ve worked for some time in this field will recognize–and have likely faced–all these issues.

No Ethics, No Sustainable Analytics Value

This uncertainty partly explains why so many companies struggle to show tangible value from their data analytics work. In our conversations with well over 100 of the Fortune 500 companies, a consistent feature was the difficulty people analytics leaders face in showing real value.

The Insight222 survey showed that 9% of companies strongly disagreed or disagreed when asked whether people analytics projects had produced tangible value over the past 12 months. In all, 30% of respondents couldn’t give an affirmative answer that their people analytics work had produced tangible value. It is one of the key reasons why we see companies restructuring their analytics functions or taking a step back from their initial commitments.

What is the solution? We believe that every company needs to have clear guidance and alignment around the employee data it should collect, especially how it should collect, analyze, store, use, and distribute it. Typically, such alignment takes the shape of a data ethics charter. However, the Insight222 survey uncovered that almost half of companies do not have any in place.

Over the past six months, Insight222 worked with 15 leading Global 500 companies (the “i222 group”) in a co-creation effort that unveiled a set of principles and led to a universal data ethics charter. The participating companies invested countless man hours, traveled to meetings in the US and Europe, and shared their expertise and know-how because they understand that to build a sustainable and successful function, both management and employees need to trust the data and analytics methods and process.

The People Analytics Data Ethics Charter

The charter developed by the i222 group and the tools to help users go from insight to action are described below. They are built on a set of six recommendations:

1. Define what’s important to you

The tenets of your approach to collecting, analyzing, and sharing people data need to be clearly defined and the people analytics team is the right team to take the lead. Other stakeholders in the organization, such as data privacy officers and the legal team, have a risk-focus when it comes to data. The people analytics team also understands the benefits that both management and the employees can derive from the proposed data analytics project and weigh risk and benefit.

2. Align key stakeholders

One of the more surprising findings of the i222’s group’s work is the sheer number of stakeholder groups: 27! Each of the 27 stakeholders has legitimate interests in the analytic work. One of the values of the co-creation is the creation of an intelligent grouping methodology to make alignment practically feasible. Once you know your key stakeholders, engage with each to assure your principles address their key concerns. If not, iterate on your principles or iterate on your messaging.

3. Demonstrate/communicate the specific individual benefit

The power of the charter is that it enables organizations to drive analytic projects with greater speed. However, this only works if the charter is well-known and accepted. A proactive and sustained communication approach is key to socializing your charter internally. The i222 group suggests including case examples where the charter was used to protect the stakeholders and create value for the employee. In fact, the i222 group’s work revealed that a key question has to be: what is in it for the employee? If no specific benefit that can be derived for employees, be careful.

4. Create a process to get to your goal

The charter creation is not a one-time conversation, a single day-long meeting, or one off-site. A truly effective charter creation is also a change management process (i.e. a multi-round conversation). This is another reason why the co-created charter is so valuable: it can speed up the initial stages, includes the credible insight from 15 highly respected, world-class organizations, and shows what best practice looks like.

5. Develop an implementation plan

Your implementation plan is adjacent to the multi-round creation process. This plan must contain clear actions steps for each phase of your people analytics project. The i222 group suggests a three-step decision model with go/no-go stage gates at each phase. These gates assure that your project stays within the boundaries of your ethical guidelines.

6. Translate your charter into action questions

A charter and decision model becomes actionable when it can be tested. Create specific questions for each stage of the analytics project to test whether the project conforms to your agreed ethical norms. Insight222 and its co-creation partners developed a set of specific, practical questions** that enable quick implementation.

Call to Arms

You will by now have noticed that taking these six steps doesn’t tackle one key problem: urgency. How do you get management to take ethically sustainable data-collection seriously before your organization has, for lack of a better term, “a Cambridge Analytica moment”?

Large organizations have a natural aversion to take on complex questions. We recommend two actions that have proven effective to create a window of opportunity: first, use the GDPR implementation. With GDPR now in place for six months, this is a great opportunity to review whether implementation is actually effective. Volunteer to take this review on and expand its remit to include data ethics. Second, if GDPR is not an option, develop a story based on your organization’s data ethic “near misses” or the (near) misses of organizations that are well regarded by senior management. Your research should point to a potential pattern and vulnerability (i.e. something that requires a systemic solution).

People analytics will only grow in importance over the coming years. Tackling data ethics now will assure that your house of ethics is built on a rock solid foundation and has lasting positive impact.

All images in this article are used with permission from the author.

*Survey is available only to Insight222 members

**If you are interested to receive these questions, contact Insight222