🚨 Datafication of Work: Tools, Ownership and AI
Navigating the new world of work requires knowing what is being used to build it: data.
Welcome! We just hit the 4,000 mark for Work3 subscribers 🎉 and I couldn’t be happier that so many of you are giving me feedback and ideas - so please make sure you reach out to me in the comments or contact me by responding to this email!
Today we’re going to cover:
The amount of data that is increasingly being captured and becoming available for both employers and employees 🛢️
How this is the key enabler for artificial intelligence applications and tool examples ⚙️
The potential benefits and drawbacks of the datafication of work for both employers and employees 👌
The ethical considerations that come with the collection and use of employee data 🚨
Let’s dive right in! 🤿
🗑️ Datafication of Work - What Data is being collected?
HR used to be considered as mostly a ‘soft’ function inside of companies, mostly about handling of relationships and more contractual reasons.
This, plus the ‘technological’ fetish that has taken the world by storm, has left it as being often perceived as a lower-value function than most others (technology, marketing etc). I personally consider this as almost ironic, always having in mind the famous proverb of the horse before the cart:
People, should be the first and foremost thought of every company. If we had the same obsession as there has been over technology, we may have avoided a lot of bad decisions. Think about it: most times this is the biggest and most delicate part of the cost structure of a company. How much of the R&D and investments go here?
At the same time, the biggest deficit in this area of ‘Human Capital’ has been the lack of data: both in quality and in quantity. The first for a good reason; it’s incredibly difficult to frame a human being in very precise categories, and there are so many variables, including emotions, that make it almost impossible to reduce all of those dimensions and be certain of causality. Lack of quantity, i.e. how many signals were recorded, was more for lack of technology and for the lagging of the industry per-se.
Forward to 2022, and these are all the types of data being recorded by companies:
Employee productivity metrics (e.g. time spent on tasks, keystrokes, mouse clicks)
Employee engagement levels (e.g. survey responses, sentiment analysis of emails and chat logs)
Performance reviews and feedback data (e.g. ratings, comments, goals)
Recruitment and hiring data (e.g. candidate sources, resume keywords, interview notes)
Diversity and inclusion metrics (e.g. demographic data, retention rates, promotion rates)
All of these data sources are then giving birth to completely new types of intelligence to companies, when collected into dashboards like these:
As you can see, there are some practical examples like:
Supply vs Demand of Skills and Roles - How many people inside the organization have a particular skill, say Python programming or how many Business Analysts are there? What’s the intersection between the two? What skills are in demand in the market for the roles the company has?
Overall Workforce Planning - What is the trend of hiring, attrition rates - both actual and forecasted according to internal data and forecasted data plugging in also external signals (CPI, etc)?
Comparison of Workers - How does Alexa Holford compare vs Mike Tyson in each
Career Development and Mentorship - How can the data help identify areas for career development and match employees with mentors who can help them grow and develop in their careers?
Employee Wellness and Engagement - How are people feeling both psychologically and physically (important for some specific jobs). I wrote a specific article on this recently.
Organizational / Social Network Analysis - Who communicates with whom, who has the most influence and how should the organization be redefined accordingly?
Financial and Compensation - How are people compensated, and how does this compare with the market? What is the impact on retention?
These’s plenty more, but you get the point.
🪙 My Data, not your Business - Blockchain and Solid
‘Data is the new oil’.
True, but this is ‘personal’ oil that is being extracted. In the age of ‘Surveillance Capitalism’ (go read that book if you haven’t) and of the ‘Attention Economy’, we use platforms that in turn use our data to run ads by us and monetize.
We give away the permissions to do that because we are locked in the usage of Social Networks, Search and so on. This is changing, because in recent years a lot of users have been starting to recognize this and act accordingly (think Apple’s privacy features, GDPR, Adblockers and so on).
Another way that is yet to fully materialize though, is the model of data ownership guided by blockchain. Blockchain technology can be used to give employees ownership of their data, allowing them to control who has access to it and how it is used. By securing employee data on a decentralized ledger, blockchain ensures that the data is not being used without the employee's consent. This can help to address the ethical concerns surrounding the use of employee data in the workplace. Plus, this data can be monetized.
Some examples for those interested:
Ocean: One of the leading blockchain-based data marketplaces. The platform provides a wealth of accessible private data that can be used for a variety of applications. Also, the Ocean Protocol staking feature allows participants to earn a passive income by providing liquidity to the Balancer automated market maker (AMM) and curating datasets. More here for those of you interested in detail.
Ceramic: a decentralized protocol for building and sharing verifiable data on the Web3 stack. It provides a secure and reliable way to create, store, and share data and applications in a decentralized manner, without relying on any centralized authority, and supports decentralized identity and provides a flexible framework for building decentralized applications.
Solid: Tim Berners-Lee, the founding father of the internet itself, has also been working away at building an alternative solution for data ownership, and it entails a decentralized and permission-based model, where users decide which data to share and which not.
I find that it’s just a question of time, nothing else, when one of these models will kickstart a snow-ball effect and take the world by storm.
Some things will still be very debatable though especially for the world of work, i.e.: who has ownership of your ‘productivity’ data at work? is it legal to be collected? if so, how much are you accountable? and so on.
🤖 The Rise of the Robots - Artificial Intelligence at Work
Now we have lots of data: this is where AI comes in.
AI can be used to identify patterns and insights that may not be immediately apparent to humans.
One example of an AI application is in the hiring process. More and more companies are using AI algorithms to scan resumes and identify top candidates based on certain criteria, namely through ATS (Applicant Tracking Systems) which end up giving something like this:
We had also discussed not too long ago how Amazon has cut back a lot on its Talent Acquisition team due to their use of AI in the hiring process, which has shown to be more efficient and effective at identifying top candidates. Just a few years before, they had scrapped it because it had showed bias against women.
Increasingly, candidates are being engaged also in the initial part of the process by chatbots (look at how many are coming into the market) who are able to answer questions, schedule interviews, and provide feedback to candidates. Chatbots can also analyze candidates' responses to questions and provide recruiters with insights into their communication skills and personality traits.
Another area where AI is being used is in performance management. AI algorithms can analyze employee productivity metrics and provide managers with recommendations for improving performance. This can include identifying tasks that are taking longer than they should or suggesting training opportunities to improve skills.
👍 OK - Is it good, or bad?
As with all things, the problem is not the tool. It’s the usage and objectives. As rightly pointed out in this article by Wired (and in a whole feature by my colleague Luca - if you read Italian check it out) this data can help exacerbate the productivity obsession that surrounds big tech. So companies will increasingly start to think that they can take decisions solely based on numbers on a dashboard, and increasingly, delegate them even to AI.
Now the good thing is that, also employees and workers are getting more and more access to information than they used to in the past thanks to this process. A good example is about salary data, which obviously reduces information asymmetry (and information is power especially in this case). It’s interesting to note that this is happening both thanks to ‘peer-to-peer’ sharing (workers sharing their salary information on sites like Glassdor, Levels.fyi and more) but also legislations are catching up (the state of New York now makes it mandatory to specify the compensation range in job descriptions).
Have a look at the level of interest in this information, according to Google Trends, and on the right, an example of how now companies are being compared very specifically around Total Compensation (i.e. base salary, stock options, perks and benefits):
All of these data sources are giving birth to a a boom in so called ‘HR Tech’: tools and startups that are looking at how to aggregate, parse, clean, and make sense of all this data:
Speaking specifically about AI, there are several concerns about biases as we’ve seen already, and o it’s important for companies to approach the datafication of work with transparency and ethical considerations, ensuring that employees are aware of how their data is being collected, used, and protected. These are some of the steps that need to be integrated:
Establishing Ethical Guidelines: Many companies are establishing ethical guidelines and principles for the development and use of AI. These guidelines can help ensure that AI tools are designed and used in a fair and unbiased manner.
Diverse Teams: Companies are creating diverse teams of developers and data scientists who work together to build AI models. This can help ensure that a variety of perspectives are represented in the development process and that potential biases are identified and addressed.
Transparency and Explainability: Companies are focusing on making AI more transparent and explainable, so that employees can better understand how decisions are being made. This can help build trust in the system and reduce the risk of unintended biases.
Regular Audits: Companies are conducting regular audits of their AI systems to identify potential biases and ensure that they are being used in a fair and unbiased manner.
Employee Feedback: Companies are actively seeking feedback from employees about their experiences with AI-powered performance management tools. This can help identify any issues or biases that may be present and inform the development of more effective and equitable systems.
Employees need to be aware and require that these are standard practices. I would expect this kind of information coming up on the company review pages (Glassdoor, etc) as it can become a really important type of information to decide whether there is a cultural match when joining.
📚 TLDR;
Datafication of work and the types of data being collected by companies, including employee productivity metrics, performance reviews, and diversity and inclusion metrics are increasingly common
The potential benefits and drawbacks, as well as the ethical considerations that come with collecting and using employee data are still unchartered territory
HR Tech is booming, with tools and startups looking to aggregate, parse, clean, and make sense of all this data.
The use of artificial intelligence in the hiring process and performance management is also becoming more common, including the use of AI algorithms to scan resumes and identify top candidates based on certain criteria and the use of chatbots to engage with candidates and analyze their responses.
There is potential to give employees ownership of their data and control over who has access to it through blockchain technology