4 min read
Earlier this week, Instructure announced that they were being acquired by a private equity firm for nearly 2 billion dollars.
Because Instructure offers a range of services, including a learning management system, this triggered the inevitable conversation: how much of the 2 billion price tag represented the value of the data?
There are multiple good threads on Twitter that cover some of these details, so I won't rehash these conversations - the timelines of Laura Gibbs, Ian Linkletter, Matt Crosslin, and Kate Bowles all have some interesting commentary on the acquisition and its implications. I recommend reading their perspectives.
My one addition to the conversation is relevant both to Instructure and educational data in general. Invariably, when people raise valid privacy concerns, defenders of what currently passes as acceptable data use say that people raising privacy concerns are placing too much emphasis on the value of the data, because the data aren't worth very much.
Before we go much further, we also need to understand what we mean when we say data in this context: data are the learning experiences of students and educators; the artifacts that they have created through their effort that track and document a range of interactions and intellectual growth. "Data" in this context are personal, emotional, and intellectual effort -- and for everyone who had to use an Instructure product, their personal, emotional, and intellectual effort have become an asset that is about to be acquired by a private equity firm.
But, to return to the claims that the data have no real value: these claims about the lack of value of the underlying data are often accompanied by long descriptions of how companies function, and even longer descriptions about where the "real" value resides (hint: in these versions, it's never the data).
Here is precisely where these arguments fall apart: if the data aren't worth anything, why do companies refuse to delete them?
We can get a clear sense of the worth of the data that companies hold by looking at the lengths they go to both obfuscate their use of this data, and the lengths that they go to hold on to it. We can see a clear example of what obfuscation looks like from this post on the Instructure blog from July of 2019. The post includes this lengthy non-answer about why Canvas doesn't support basic user agency in the form of an opt out:
What can I say to people at my institution who are asking for an "opt-out" for use of their data?
When it comes to user-generated Canvas data, we talk about the fact that there are multiple data stewards who are accountable to their mission, their role, and those they serve. Students and faculty have a trust relationship with their educational institutions, and institutions rely on data in order to deliver on the promise of higher education. Similarly, Instructure is committed to being a good partner in the advancement of education, which means ensuring our client institutions are empowered to use data appropriately. Institutions who have access to data about individuals are responsible to not misuse, sell, or lose the data. As an agent of the institution, we hold ourselves to that same standard.
Related to this conversation, when we hear companies talking about developing artificial intelligence (AI) or machine learning (ML) to develop or improve their product, they are describing a process that requires significant amounts of data to start the process, and significant amounts of new/additional data to continue to develop the product.
But for all the companies, and the paid and unpaid defenders of these companies: you claim that the data have no value while simultaneously refusing to delete the data -- or to even allow a level of visibility into or learner control over how their data are used.
If -- as you claim -- the data have no value, then delete them.