Data is at the center of business strategy today. We use data to make decisions, define who we are (the quantified self and the data-driven enterprise) and even to predict the future. So it is not surprising that fights over data ownership are becoming more common.

This dispute is coming to a talent management system near you. In most companies it is the company and not the individual who owns the data about their skills, learning and work experience. And many HR administrators will say of course, how else could it be?

Their employees have already shown them the future though. Professional social networks like LinkedIn have better data about people, there current activities and interests, and even what their skills are, than do proprietary in-house systems. Go to LinkedIn, search for your company and you can find most of the people who work there. You can see what groups they have joined and how people in their networks have endorsed their skills; where they went to school, where they have worked, and where they volunteer. Now try to find this information from your internal systems. People are volunteering to put data into systems like LinkedIn because it becomes their data (or they think it does) and because it provides them with value.

But there are problems with the LinkedIn approach. Who really owns the data? The user agreement covers this of course. In section 3 it says clearly that you own your own data. That sounds pretty reasonable. But think this through. Under 3.1.1 you can delete content. Is that the only right one should have? And just what is content? In today’s world, ‘content’ is a lot more than just text and images. It includes the metadata, which LinkedIn gives itself the right to remove or modify, and even the relations between people. Is your list of connections on LinkedIn ‘content’? How about all of the updates you have posted to LinkedIn or your comments in Groups? A lot more public discussion of what is included in ‘content’ and what it means to ‘own’ content.

Let’s remember we are talking about digital content and not a physical object. The interesting thing about digital content is that many people can own it and it is easy to modify. The Creative Commons movement has thought hard about this and has come up with a number of different ways to license content. But even Creative Commons struggles with shared ownership, and it is shared ownership that we need for HR data.

Shared or collaborative data has some important properties:

  1. It is created by more than one person or organization, each of which establishes some rights
  2. It often changes over time
  3. The nature of a person or organization’s rights can change as the relationship between them changes

One can develop these properties into a set of rights that define what actual data ownership means:

  1. The right to delete data – this right can only be exercised if it does NOT violate other people’s rights who share ownership of the data
  2. The right to edit or modify data – this right depends on agreement between the people who created the data, the best way to manage this is the wiki approach where each version is stored, edits can be rolled back, or versions can be branched
  3. The right to add to or comment on data – this generally establishes some level of ownership over the data concerned (this is the surprising conclusion)
  4. The right to control who can view the data – other than the people who have established a right
  5. The right to view data – but not to modify or comment on it or even to share it

Any agreement on data ownership should address these rights and say how they can be exercised.

We are talking about data, strings of zeros and ones. Data is only usable inside a software application. This means two things. First, there needs to be a way to transfer the data from one system to another while respecting shared ownership rights. In some cases this will be very difficult. So the other issue comes to the fore, access to data through APIs. If I own the data in any meaningful sense I should be able to view it through the software platform of my choice. If this right becomes generally accepted it will have a profound impact on how SaaS companies do business.

What does all this mean for the ownership of HR data?

  1. Companies, individuals and teammates all need to share ownership
  2. There need to be controls on who can modify and view data

Talent management needs to become more about the talent. About giving people the power to manage their careers. Too many of today’s systems are either company-centric or platform centric. The individual and the people they work with on teams get no respect.  So what are the options?

Company Owns Doesn’t work. Individuals have to be coerced into providing data, the value of the data is reduced as it is incomplete (it does not include critical data that previous employers or the individual owns), collaboration with outside people and partners is made difficult. Individuals move over to external services (like LinkedIn) to manage their own data.

Platform Owns Doesn’t work. Both companies and individuals lose de facto ownership of the data. They cannot effectively exercise their rights. Companies try, usually ineffectively, to prevent employees from entering data on public platforms, individuals cannot put all of the relevant data into these platforms without violating employment agreements, privacy and security are easily compromised. This is the actual status of data on systems like LinkedIn.

Individual Owns Doesn’t quite work. This one gets closest to the mark, but most work today is done in teams and for one or more companies. Teammates have a legitimate ownership stake. So do companies. Platforms that say the user owns the data are not clear about what that means or how they allow people to exercise their rights.

The current situation prevents both individuals and companies from realizing the full value of HR data. What is the solution?

Shared ownership. With the rights of that ownership clearly set out. And the ability to exercise those rights coded into software platforms and data formats. That is what we are working towards at TeamFit. How we are doing this will be the theme of my next post.