Skip to Content

How to “Go FAIR”: The Key Business Decisions to be Made

Natalie Stanford
4 Jul 2022

There are many ways to “go FAIR”, choosing the right approach for your organization is critical for success.

The FAIR Guiding Principles detail characteristics of data that support better data management. They stand for Findable, Accessible, Interoperable and Reusable.

When implementing FAIR in your organization, there is no “right way” to do it. You approach will depend on factors including: current data management maturity; what data you manage; company culture; your budget, time, and skills, among others. Here are things you should consider during your journey of “going FAIR”.

Enabling Findability

Findability is about ensuring that data can be searched, identified, and retrieved by a range of users. Enabling Findability involves getting the right balance of technology, processes, expertise, and data management support. You need to consider the following: 

  • Metadata development and capture: you need to ensure that all data is labelled with enough key attributes that your users can easily search for it. For instance a user might want all data collected within the last 12 months that relates to Breast Cancer. Therefore, you would need to introduce key attributes such as Date of Collection, and Disease Type.
  • Storage / Centralisation of data: for data to be physically findable it needs to be ‘centralised’, so your users know where to look. You could achieve this through physical centralisation e.g. a data lake, or a hub and spoke style centralisation where data is kept in domain specific storage, and a catalogue us used to direct users to its location. What makes sense for your business will depend on your tech and process history, and future expectations.
  • Searching for data: you need to enable effective querying of metadata. Solutions such as cataloguing software supports key-word searches and filtering of data, to support identification of data and location.

Enabling Accessibility

Accessibility is about making the right data available to the right people, at the right time, with the right supporting information. This is not a free-for-all, and data access should be restricted as much as needed (e.g. privacy, sensitivity etc.) and as open as allowed (e.g. licensing, regulations etc.). You will need to consider the following:

  • Data Vs Metadata: FAIR makes a key distinction between metadata and data. You will need to decide what information about the data to expose to ensure staff can identify suitable access permission channels as needed.
  • Legal and contractual requirements: what legal or contractual restrictions does your data have?  You need to ensure that users are only able to view and access data that they have adequate permissions for, which also means you need to consider how to match users to permissions.
  • Technical considerations: What is your current technology stack? FAIR emphasizes making data accessible through a standardized communication protocol, for instance through a secure internal web app.

Enabling Interoperability

Interoperability is the ability of people, computer systems and/or software to exchange and use information. You then need to decide what levels of interoperability are both feasible and beneficial to your use cases over your expected timescales to ensure you get the most value from what you implement.

There are varying degrees of Interoperability, each incrementally improving the holistic interoperation of information: 

  • Foundational Interoperability: this involves business information moving between different systems in a way that the purpose and meaning of the data is preserved e.g., a report in Word, or a PDF. In this instance it is the user that interprets the understanding; the systems only support the transfer of the file structure. 
  • Structural Interoperability: this involves business information being exchanged between systems, with the systems able to absorb the data automatically using pre-defined context mapping of data columns. In this instance the systems don’t ‘understand’ the data, simply implement programmed mapping.
  • Semantic interoperability: this involves systems exchanging information seamlessly through full understanding of the origin, context, and meaning of the data. It relies on structured Formats (e.g. JSON, SML, RDF, API, etc.), and Controlled Vocabularies, Ontologies and Synonym libraries to ensure that each system can seamlessly interpret the true meaning of each piece of information.

Enabling Reusability

Formulating the right data management approach to Findability, Accessibility, and Interoperability, are the foundations of Reusable data. In addition, to ensure Reusability you need to ensure you have the right:

  • Data Linking / Provenance: for data to be truly reusable a user needs to understand the original context of the data, including processing methods or transformations that were performed to achieve the current data. This is particularly important in scientific data. The origins are the key to valid reuse.
  • Technology: data that relies on specific and / or legacy software to interpret and understand needs supporting technology to ensure suitable re-use. This includes e.g. containerization, virtual machines, or legacy computer set-ups to support older data.
  • Culture: for effective re-use, culture is king. Through establishing good sharing practices, suitable organizational support and structures, and creating incentives for FAIR data sharing, a company can rapidly improve the reusability of its data.

Conclusion

Strategically implementing the FAIR Guiding Principles enables you to establish a data-drive organization, impacting many aspects of your business value, including: enabling analytics capabilities; improving the speed and reliability of auditing; and ensuring long term value of laboratory data. If you are unsure where to start, Capgemini’s many years of experience with FAIR and data management will help you identify the best strategies for you, and enable you to truly embrace becoming a data-driven R&D organization

Meet our expert

Natalie Stanford

Life Sciences and Data Management Consultant, Capgemini Engineering
Natalie is a Chemist, Biologist and a FAIR data expert. She has been supporting clients to achieve their R&D data management goals since 2013, delivering support, strategy, and roadmaps to clients across academia, Fast Moving Consumer Goods, and Big Pharma.