The corporate value of data is well-understood. Where organizations often struggle is in leveraging that data[1]. To get value from the vast amounts of data that exists within an organization, businesses need to design a robust data strategy supported by data architecture, which in turn requires a suitable information technology architecture[2],[3]. The data strategy is supported by a data management program through defined procedures, roles, and responsibilities1,[4].
Understanding the importance of a data strategy is a first step. The next is to consider the approach. There are two opposing approaches to the data strategy: a defensive and an offensive one2,[5]. An offensive data strategy focuses on a business value, such as revenue or customer satisfaction. Typically, this involves quickly leveraging data for particular use cases such as predictive modeling5. In our opinion, this strategy would seem to be a good match for pharmaceutical research, where time to market is a crucial factor[6]. It allows for the flexibility that the industry requires and allows companies to quickly gain insights from the data to support the next iteration of discovery and lead optimization[7].
A defensive data strategy means avoiding negative outcomes, being compliant, limiting fraud, or preventing theft2,5. In pharmaceutical development, the dominant factor is regulatory compliance, and data is assessed based on this requirement. Compliance in the industry needs to consider several factors, including harmonization, standardization, data quality, security, privacy, and avoiding ambiguity. Artificial intelligence and machine learning use cases also require large data sets, high-quality data and analytics experts, now more readily available thanks to early adopters and advanced tools, applications and services6. In our opinion, a defensive approach would also be a good fit for the pharmaceutical industry.
Data Preservation
There is another factor that should be considered, at least for older, mature pharmaceutical companies. Due to the need for intellectual property protection and the regulatory requirement to carefully store raw data, many companies started their digital transformation early and now have many years of data to leverage. Generally, scientific results do not lose their validity over time; therefore, it is worthwhile, and important, to retain this data. The challenge, however, is that historically companies have often captured data in various systems that were decommissioned at the end of their lifecycle and replaced[8],[9]. This is what we often observe in our role as a consultant and system integrator for pharmaceutical companies. Current and legacy data coexists, but might have different data structures, using different semantics, very often distributed in silos, systems or even sites across the organization. As such, the data are distributed and heterogenous, a situation that is likely to continue due to mergers and acquisitions and the limited lifecycle of applications9,10.
When applications reach the end of their lifecycle, a centralized place to conserve and leverage the legacy data in long term is needed. In my experience, any consolidation of current and legacy requires always some level of harmonization and standardization.
Deciding on the best approach
There are many data strategies or data architecture concepts published in the literature and applied in practice. Examples include data mesh, data fabric, data warehouse, data lakehouse, data vault, and data cloud[10].
Pharmaceutical companies depend on flexibility and quick data availability during research, and on data accuracy and compliance in later stages, as well as the option to benefit from older legacy data. Therefore, it is difficult to state a good data strategy that might be a good match for every pharmaceutical company.
In our opinion, each company has its own specific needs and criteria when considering a data strategy. This can be the existing IT landscape, existing roles and responsibilities, procedures or even the corporate culture. Moreover, it is of utmost importance that the data strategy describes how corporate data should support the business strategy. A data strategy is, therefore, as specific as a business strategy.
To derive a suitable data strategy, we recommend several key steps:
- Start with your business strategy, identify related key objectives for leveraging the data. Think about use cases and their business value and priority. This information will help you to gain clarity about the requirements for the data strategy and potential changes.
- Design your data architecture and IT architecture according to key objectives and priorities[11]
- Establish or adjust your data governance framework accordingly4
- There is the option to have different data strategies for different domains (research, development, clinical). However, different domains should not be isolated. So, while there are use cases where data from different domains are needed, e.g. translational medicine, a common semantics is recommended.
- Think long-term. What happens with your data if systems are replaced? 1
Planning the strategy
It is not advisable to pick a predefined data strategy but rather follow a well-considered approach to derive your data strategy. Start with your business objectives and how those correspond to your data strategy. Get clarity on data consumption use cases, data requirements and apply priorities. If needed, define specific domain strategies for research, development, clinical, manufacturing. With this approach it is possible to develop a tailored data strategy that supports the business goals.
About the author:
Christian Ikier is Associate Director at PharmaLex, supporting life sciences companies with their digital transformation and helping to make scientific data more available, findable and manageable. A chemist by training, Christian has spent his career in IT in the life sciences, including more than 10 years as a consultant at Osthus, now part of Cencora.
[1] Why Becoming a Data-Driven Organization Is So Hard, Harvard Business Review. https://hbr.org/2022/02/why-becoming-a-data-driven-organization-is-so-hard
[2] What’s Your Data Strategy? Harvard Business Review. https://hbr.org/2017/05/whats-your-data-strategy
https://hbr.org/2023/06/your-data-strategy-needs-to-include-everyone
[3] How to build a data architecture to drive innovation—today and tomorrow, McKinsey. https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/how-to-build-a-data-architecture-to-drive-innovation-today-and-tomorrow
[4] Mahanti, R. (2021). Strategy and Data Governance. In: Data Governance Success. Springer, Singapore. https://doi.org/10.1007/978-981-16-5086-4_3
[5] The O-Line, Deloitte. https://www2.deloitte.com/us/en/insights/industry/public-sector/chief-data-officer-government-playbook/2023/offensive-data-strategy-in-government.html
[6] Russo A., Falk B., Sinha N., Dannacher F.,Intelligent biopharma. Forging the links across the value chain
[7] Guo M, Wang Y, Yang Q, Li R, Zhao Y, Li C, Zhu M, Cui Y, Jiang X, Sheng S, Li Q, Gao R. Normal Workflow and Key Strategies for Data Cleaning Toward Real-World Data: Viewpoint. Interact J Med Res. 2023 Sep 21;12:e44310. doi: 10.2196/44310. PMID: 37733421; PMCID: PMC10557005
[8] Data silos threaten efficiency levels for nearly half of pharma businesses
[9] 2015 Laboratory Data Knowledge Management Report
https://www.pharma-iq.com/informatics/whitepapers/2015-laboratory-data-knowledge-management-report
[10] Knight M., Nov 2023, What Is Data Architecture? Components and Uses
https://www.dataversity.net/what-is-data-architecture/
[11] Developing a Robust Data Strategy: Navigating the Digital Era, Dr. Fatih Nayebi. https://www.linkedin.com/pulse/developing-robust-data-strategy-navigating-digital-nayebi-ph-d–n5kje/
