InBloom: Data Driven Learning Flounders Amidst Panicked Political Climate

“InBloom is one of the first major big-data casualties – a victim of exaggerated fears and misunderstanding.” -The Economist

InBloom is a story of unfulfilled data potential. The narrative is perhaps a familiar one … U.S. political forces grappled with the trade-offs between privacy and efficacy, and the confused regulatory environment that resulted hindered important, life-improving innovation.

This particular story takes place in the education sector. InBloom was an education non-profit endowed with $100 million from the Gates and Carnegie Foundations. The organization’s mission was to consolidate and clean student data from across the many fragmented student data warehouses. The ultimate goal was to make the data available for proprietary InBloom  and district-approved third party dashboards that would 1) allow teachers to tailor instruction to each student and 2) would allow administrators to make more informed management decisions. Many initially praised InBloom’s attempt to make student data more actionable. However, the organization faced a tsunami of well-mobilized parent and legislative resistance and was forced to wind down in 2014.[1] Heres’ the value creation/value capture story as well as ideas on how future education companies can avoid InBloom’s fate.

Value Creation – The Promise of Data Driven Education

Let me start with the basics … The current system of education is outdated. Students that can’t keep up with the average classroom pace fall through the cracks, while advanced students grow bored and restless. Research and common sense consistently indicate that people learn in different ways – the preferred mode of learning, the desired nature of content, and the pacing can vary dramatically from student to student. Data has the potential to play a huge role in in delivering a more student-centric form of learning. Here are just a few of the various uses cases through which data can create value in the classroom:

  • Provide teachers with regular progress reports on each individual student so teachers can allocate their time accordingly
  • Flag students that are late or absent[2]
  • Use predictive analytic models to identify students who are at risk of not graduating[3]
  • Spur parent engagement via regular student progress reports to guardians[4]
  • Leverage demographic data to identify and direct resources to those schools that are serving students with the greatest needs
  • Identify and automatically push content to students depending on their current level of mastery and individual learning roadblocks
  • Provide data around efficacy of technology tools to inform the procurement process

And these examples constitute just a drop in the bucket. McKinsey’s education team estimates that the use of student data could “unlock between $900 billion and $1.2 trillion global economic value; upwards of $300 billion would come from improved instruction.”[5]

Value Capture – Multiple Paths to Sustainability

InBloom was a non-profit because they knew any profit motive would invite intense public scrutiny. Also, capturing value in the K-12 sector is difficult – budgets are limited and sales cycles are long. Yet, there are a number of education data companies that have found creative ways to capture the value they create. A few business models that have been successful include:

  • Create a marketplace for third party data related tools and charge associated transaction fees (Amazon education)
  • Charge other data management developers an integration fee for plugging into the system (Clever model)
  • Provide base product for free and upsell premium features (Newsela model)

What Went Wrong? …  The Political Obstacles to the use of Data

So what happened? InBloom underestimated the anti-surveillance, pro-privacy political winds mobilizing against them. In a post-Snowden era, concerns around privacy are entirely valid, but I believe they need to be balanced with an appreciation for how data can also enhance our lives. Specific parent/teacher concerns included the following:[6]

  • Student data being sold to advertisers for a profit
  • The use of data to track or restrict students into certain academic paths (e.g. labeling a student underperforming.)
  • Data breaches and identity theft
  • An over reliance or quantification of learning
  • Teacher anxiety that their jobs are being automated and that data will fundamentally change the nature of their work (e.g. challenges to their professional identity)

The political obstacles to the collection and utilization of learning data are formidable. A 2016 report from the TRUSTe/National Cyber Security Alliance (NCSA) Consumer Privacy Index indicated that “more Americans are worried about their data privacy than they are about losing their main source of income.”[7] As such, compromise will be required on all ends – parents will need to make allowances, policy makers will need to craft responsible and flexible regulation, and companies will need to assuage concerns through a variety of measures. McKinsey suggests that data education companies can learn a lot from other sectors. Some recommended tactics include developing unambiguous and accessible privacy policies, more effectively communicating the benefits of data, adopting sector wide data standards, and treating districts as partners in the communication process.

As the Economist wrote, InBloom was one of the “first major big-data casualties – a victim of exaggerated fears and misunderstanding.”[9] While I’m sure it won’t be the last, education data enthusiasts have learned from the past and appear to be moving forward with renewed optimism.

 

[1]http://blogs.edweek.org/edweek/DigitalEducation/2014/04/inbloom_to_shut_down_amid_growing_data_privacy_concerns.html

[2] Interview with David Roth, Senior Direct of Data Systems at Aspire Public Schools

[3] Interview with David Roth, Senior Direct of Data Systems at Aspire Public Schools

[4] http://www.mckinsey.com/industries/public-sector/our-insights/protecting-student-data-in-a-digital-world

[5] http://www.mckinsey.com/industries/public-sector/our-insights/protecting-student-data-in-a-digital-world

[6] https://www.parentmap.com/article/student-data

[7] http://www.cbsnews.com/news/truste-survey-more-americans-concerned-about-data-privacy-than-losing-income/

[8] http://www.economist.com/blogs/schumpeter/2014/04/big-data-and-education

Previous:

Amino: Taking the Guesswork out of Healthcare

Next:

Broadway Enters the ‘Big Data’ Era

2 thoughts on “InBloom: Data Driven Learning Flounders Amidst Panicked Political Climate

  1. Kyla,
    This is fascinating! I’m obviously biased in favor of data-driven initiatives across the board, and definitely am not shy about surrendering personal data if I think the benefit will outweigh the risk. But my in-laws, for example, would have been terrified of InBloom’s data play and the data security issues it presented.

    It seems, however, that there are ways to make people more comfortable with data gathering. I wonder what strategies InBloom could have employed to ensure relevant groups that the data they gathered and utilized would be managed responsibly, securely, and then used inclusively. How could they have proven to core constituencies that the benefits of data outweighed the risks in this case?

    1. Thanks for the thoughts, Sonali!

      While some of the outrage directed at InBloom was, in my opinion, unwarranted and paranoid, there are definitely things InBloom could have and should have done differently to address concerns. This was definitely still a failure on the part of InBloom. In terms of your point above, there should have been a lot more education to parents about the use cases of the data. That type of messaging was directed at the schools who were deciding whether or not to use InBloom, but InBloom should have recognized that parents also have a voice in the decision process. If parents understood exactly how the data would be used in the classroom and how it would improve learning for their child, they may have been more open to it. Other tactics mentioned above include developing unambiguous and accessible privacy policies, galvanizing a number of different edtech companies to adopt sector wide data standards, and treating districts as partners in the communication process (e.g. working with districts to make sure parents’ concerns are addressed).

Leave a comment