What is Data Liberation

And why should I care?


I believe we all benefit from making the world more open.

Businesses like to tie users to themselves. One way is to make it hard for the users to change.

Naturally, this model benefits businesses. Naturally, it hurts the end users.

With everything moving to the cloud and businesses become more about services, not goods, this trend is reversing.

The cycles are shorter. Customers became more flexible. They are more demanding. They are more willing to change. They are easier to win for businesses — and easier to lose as well.

These days, when it comes to Internet services, the above translates into the whole new paradigm:

Don’t tie users to the service. Instead, give them a reason to keep coming back.

It can be painful to move between services that constitute a large chunk of one’s live.

But, ultimately, the services that do focus on giving people the flexibility they want will take over, while those services, that fight hard to not let people go, will get extinct.

This is why we all should understand what Data Liberation is.


Data liberation is the natural step of understanding that in the long run:
  • Successful companies create products that offer all-round end user experience.
  • External services that leverage one’s data are “promoted’ to “end users” and treated accordingly.
  • The services that “sit on their data” will be and already are losing.
On the technology level it translates to:
  • Design internal data access APIs in a robust way assuming that they may be exposed externally some time soon.
  • Specifically, design for scalability, SLAs, permissions control and backwards compatibility.
  • Whenever possible, use external-grade APIs to access internal data.
  • Focus more on testing the APIs instead of testing internal data representations.
  • When allowing access to user data, focus early on making sure that the users’ privacy and security are not at risk.
Community-wise it translates to:
  • Publicly declare company’s policy on data.
  • If certain data access mechanisms can potentially change the landscape: present designs to the public, give public talks, be open to feedback and suggestions.
  • Open source data-driven pieces of code, canonical example usecases, maybe even the tests.
  • Make sure the team has experienced people whose job is to verify that the technology provided by the company goes in line with the world’s understanding of data liberation and adjust the path accordingly.
  • Make sure the necessary data-related legal precautions are taken. The goal is to be open for long-term external partnerships, not to make it easy and/or beneficial for a third party service to “crawl” one’s data and use it somehow.

This is my second, after Open your Source, post about making the world more open technology-wise.

Please let me know if in the future you would like to read more about data technologies and how do they make our lives better.