For years, leaders have tackled the question of, “How transparent is too transparent?” One of the most common debates, for example, is whether to share salary information across an organization. This debate will rumble on as employee data is increasingly used. To ensure ethical data collection and use, it’s vital that businesses do not assume unbarred access to all employee data. Especially as increasing amounts of sensitive data are utilized (e.g., behavioral and health data), two rules should stay front-of-mind:
- Aggregate when possible.
Individual worker data should be used only on rare occasions when there’s a real business case for it. The business case should also be clearly explained to employees for why individual data is being used. For example, individualized data would be vital for assigning internal resources to a new project, as tapping into individual skills profiles would lead to greater efficiency in allocating resources. In general, however, aggregated data should be the norm. For example, behavioral data measuring employee concentration levels would only be ethically collected in the aggregate to determine, for example, whether certain locations in the office foster better focus levels or the impact of the latest restructuring announcement.
- Don’t collect actual content.
When organizations collect communication-related data – through email, Skype and other personal conversations (virtual and physical) – a basic principle is to never analyze what people are actually discussing. Understanding collaboration across a team requires knowing who is speaking to whom and how often, as well as behavioral measures such as respect or attention. What’s actually being said is of little value. Failure to get the balance right exposes organizations to a breakdown of trust, suspicion from the workforce and even public scrutiny and reputational damage.
This rule is particularly pertinent for one area of Talent Intelligence: Organisational Network Analysis (ONA). While AI is great at “the science of the job” (data analytics and pattern recognition), people are great at “the art of the job;” for example, visual cues, emotion, empathy, judgment, and social context. ONA helps to quantify these qualities by analyzing the relationships and networks employees build.
Solution Spotlight: Riff Learning Platform
We’ve been speaking to Beth Porter at Riff about their video-conferencing solution, which conducts a type of ONA, taking data generated through their audio-visual tool and applying machine learning to generate insights about network activity and the strength of engagements within and across teams.
We took a closer look at what data is being collected through the Riff Platform video conference tool, and we ask how it’s being collected and why:
- Vocal data (volume and tone only). No words are captured nor do recordings get preserved after insights are derived from them.
- Facial/gestural data. Activity in the call is matched to well-known visual patterns, such as nodding, smiling, or averting attention to identify areas of agreement, discord, engagement, and inattention. Specifically, we are most interested in conversational markers such as “affirmations” which are small vocal or gestural signals of attention and encouragement. Again, recordings are discarded after insights are derived.
We didn’t beat around the bush, asking Beth how likely it is for the employee to reject this kind of monitoring. She responded, ‘As we have already seen through our experimental research work at MIT, people are willing to be measured in their activities as long as there is an upside for them personally.’ This goes back to the first rule of ethical talent intelligence – clearly and transparently communicating the give-to-get ratio.
Beth went on to say, ‘In our work with learning experience providers, for example, people are using the video tool to enable better online communications and the measurements are reflected back to users in the form of immediately usable feedback. People who make effective use of the tool and this feedback perform better on assignments requiring collaboration.’
Your employees are not rats in a lab – don’t treat them as subjects of a data analytics experiment. Aggregate where possible and don’t collect content to respect individual’s privacy and maintain trust. Ensure there is mutual benefit and communicate that mutuality clearly and consistently to succeed in Talent Intelligence efforts.
This blog is part of a series on Talent Intelligence and the Big Brother Burden.
In 2018 there was plenty of excitement surrounding the potential of People Analytics. We published our take in Talent Intelligence: Unlocking People Data to Redefine How Humans Need to Work. We believe that Talent Intelligence will be the secret to solving your biggest talent crises:
- Finding and retaining top talent
- Fostering productivity, performance and well-being; and diversity and inclusion
- Driving agile, flexible attitudes toward human-to-human and human-to-machine collaboration that unlocks innovation
But before you can reap the rewards, there’s one big hurdle standing in your way: the Big Brother Burden. We break down the Big Brother Burden into four golden rules for the ethical collection of employee data. Organizations have to play by these rules if they want Talent Intelligence efforts to stick.
In this blog series I’ll take you through the four golden rules in a bit more detail…
- Introducing The New Rules
- The New Ts&Cs: From Terms & Conditions to Transparency & Clarity. Make the “give-to-get” ratio clear and compelling.
- Remember Who Owns the Data. Portability counts. Employees own their data and have a right to download it and take it with them.
- Who’s Watching? Individual worker data should rarely be used and only when there’s a real business case for it. Otherwise, aggregate data to ensure anonymity.
- Only Ever If They Opt-In. Taking part in data collection in the workplace must always be optional, not mandated.
Dive in to get your data ducks in a row. Only then can you make the most of Talent Intelligence tactics.