Level Up to Strategic Data Sharing
Your data assets are key to developing new value for your customers and giving you clout in digital ecosystems.
Topics
![](https://sloanreview.mit.edu/wp-content/uploads/2023/02/2023SPR-Wixom-1290x860-1.jpg)
Jon Krause/theispot.com
Most organizations have learned how to share data tactically. They’ve put technical infrastructure and processes in place so that they can easily transfer data when necessary to comply with regulations, execute transactions, or provide a service. But far fewer companies have begun to explore the opportunities created by a more strategic approach to data sharing.
In today’s digital economy, business leaders need a data-sharing strategy to pursue novel digital solutions — and move in new business directions. Strategic data-sharing practices allow an organization to share data quickly and with control for specific opportunities with specific partners, and to share it repeatedly, with an emphasis on value creation. As strategic data-sharing practices mature over time, purposeful, fast, and creative data sharing becomes the basis for business model innovation and the resulting payoffs.
Get Updates on Leading With AI and Data
Get monthly insights on how artificial intelligence impacts your organization and what it means for your company and customers.
Please enter a valid email address
Thank you for signing up
Energy management and industrial automation company Schneider Electric is a case in point. It uses its internet of things platform to share data with its business customers in order to discover and deliver energy management solutions.1 In 2016, company leaders saw IoT as a significant opportunity to develop new products and services based on the monitoring and analysis of connected assets. Along with the data-sharing platform, they developed a tool to collect both structured and unstructured data about the company’s products in use at a customer site, drawing on sensor data and maintenance logs. Using a dynamic risk monitoring algorithm, the tool detects operating risks and their sources, and it allows facility managers who purchase the product to manage the performance of assets themselves or use a Schneider Electric service center.
While most pilot sites kept the data and its analysis within their internal systems, early adopters who were willing to share data in the cloud got more value through features that derived additional insights by analyzing data from multiple pilot sites and/or customers.2 For example, Schneider Electric has reported that customer Hilton Hotels & Resorts has achieved 14.5% energy savings since 2009, based on monitoring real-time consumption and pricing data — a key outcome, given that energy is Hilton’s second-largest operational cost after labor.3 By 2021, Schneider Electric’s business related to digital solutions and the EcoStruxure IT platform (comprising software, controls, connected products, and services) generated more than 50% of the company’s total revenue.
Strategic value can also arise from sharing data within an innovation ecosystem. Health insurance provider Elevance Health (formerly Anthem) shares data with academics, researchers, and other partners to identify ways for AI models to inform complex health care problems. In 2019, Anthem’s innovation team assembled 12 years of de-identified medical claims, prescription claims, and lab data associated with 45 million individuals. Then the company set up a data-sharing platform to provide sandbox development access to the curated data sets for AI model training. By August 2021, nearly 50 vetted partner organizations comprising more than 600 developers and researchers had generated over 30 projects addressing wide-ranging health care problems. They included improving the management of diabetes, predicting breast cancer risk, treating heart disease, and reducing avoidable hospital readmissions. On average, about five projects made it into production each year from 2019 to 2021. C19 Explorer, a project launched in 2020 after two months of sandbox development, became one of Anthem’s most-used COVID-19 resources. The sandbox has been moved to Elevance’s Carelon service subsidiary and is now branded as Carelon Digital Platforms.
The examples of Schneider Electric and Elevance Health illustrate the benefits of having strategic data-sharing practices that help organizations cocreate with customers, engage in novel partnerships, and pursue ecosystem opportunities. Since 2018, we have studied digital initiatives at 30 organizations that were engaged in strategic data sharing. (See “The Research.”) We identified four key practice areas that these organizations focus on to advance this capability:
- They invest in data management practices that increase data liquidity, by identifying strategic data assets and making them usable, available, and combinable.
- They reduce friction via service platforms and oversight practices that facilitate the fluid exchange of data assets, by implementing flexible controls and transitioning from bureaucratic, manual governance approaches to automated, repeatable oversight processes.
- They set up win-win relationships with partners, using policies and processes to enable trust and collaboration even in the face of power differentials.
- They proactively manage innovation initiatives as value-realizing projects, set specific revenue goals, and assign them commensurate resources and managerial attention.
Let’s look at each of these areas in more detail.
Increase Data Liquidity
Data assets must be easy to reuse and recombine.4 Highly liquid data assets are accurate, standardized, secured, and generally fit for use. Organizations create highly liquid data assets by drawing on data management practices — foundational ones like master data management and data quality management, intermediate practices like data standardization and data integration, and advanced ones like using synthetic data and data obtained from external sources. Because these practices require time, money, and skills, organizations should apply them to data assets with strategic value.
PepsiCo has shared its highly liquid data assets about product sales and niche consumer needs with its retail customers.5 These data-sharing efforts were built on a foundation established in 2012, when the company adopted data management practices for cleansing, standardizing, and consolidating data from across its business silos. Its global IT unit established a data taxonomy that reconciled the different product names and formats used by different country operations around the world; this allowed data from global operations to be mapped into a single report on worldwide product sales. The IT unit had also adopted master data management practices and housed product sales data in a single enterprise data warehouse, later modernized into a cloud-based data lake. Finally, it established data integration routines and tools that offered the ability to tie PepsiCo product sales to other data assets, including the company’s consumer data asset, which covers anonymized shopper behavior on 110 million U.S. households. This service gave retailers insight into demand for PepsiCo’s products or particular categories, even at specific stores, and it suggested sales and marketing improvements. Today, PepsiCo uses these data assets to fuel a data service for its retailers.
Even companies in highly regulated environments can make their data more readily accessible. Elevance Health’s team worked for close to a year to develop data management practices that would help the company increase data liquidity while meeting data privacy and security requirements that would properly secure its patient data assets.6 The company had to protect the regulated health data from being used in ways that could unintentionally compromise privacy or reveal sensitive information. It began by de-identifying personal, confidential, and proprietary data as required by law. For example, the team altered birth dates and used only the first two digits of postal codes for member and claims data. It replaced sensitive data with tokens — random strings of meaningless data — so it would be useless in the wrong hands. It also created and used synthetic data when possible.7
Next, the Elevance team worked to ensure that the data assets could be understood by and relevant for innovation partners. It created a public data dictionary that described the technical and semantic characteristics of the data assets.8 It also tested the data assets and adjusted them to be more useful for AI model training. To do this, it drew on nearly 4,000 different predictions about its consumers’ health based on the complete data held internally and then used those findings to benchmark the accuracy with which the new data sets could make similar predictions. Once the internal users were satisfied with the predictive power of the digital sandbox data assets, the team felt ready to offer the assets for strategic data sharing.
Reduce the Friction That Slows Data Sharing
Organizations must reduce friction associated with gaining permission and data-provision logistics if data of strategic value is to be shared in a timely way. In most organizations, data sharing involves completing forms, waiting for someone to approve the request, and working with IT to add a user to a system and set them up with new software and training. Such friction can act as a check on unacceptable data use, but when it comes to strategic data sharing, friction slows discovery and value creation. Organizations with successful strategic data-sharing initiatives exploit flexible controls and advanced technology to transition from bureaucratic, manual governance approaches to automated, repeatable oversight processes.
Organizations must reduce friction associated with gaining permission and data-provision logistics if data of strategic value is to be shared in a timely way.
Innovation team leaders at Australian financial services company ANZ envisioned a highly governed, secure place where vetted partners could perform analyses on de-identified banking data assets. The team wanted to reduce friction by creating a standard framework that could be used to onboard data-sharing partners in a consistent and efficient manner. With help from a third-party provider, ANZ identified contractual parameters that could be adjusted to manage distinct obligations regarding development standards, intellectual property rights, and data use. It then used the provider’s secure technical platform to control data access for partners after contracts were set up. ANZ built out a process with developmental stage gates (such as time-bound access, customizable hardware setups, data curation plans, and approvals) to allow for adjustments and ensure proper oversight as a project unfolded and new data needs arose. The combination of parameterized contracting, a dedicated secure data exchange platform, and a stage-gate process allowed ANZ to create a repeatable onboarding process despite variations in data-sharing partners and purposes.
Create a Level Playing Field
Organizations adopt policies and processes for collaborating with partners so that they can cocreate novel digital solutions even in the face of power differentials. To do this, they track shared value, create ways to collaborate that allow people to solve problems on the fly, gain a comprehensive understanding of stakeholder needs, treat partners like customers, and prioritize clarity when developing joint goals. All these practices helped the organizations we studied establish trust with partners and generate excitement about future opportunities.
Access to PepsiCo’s strategic data assets appealed to retailers because it promised visibility into shopper behavior on a broader scale.9 The company evaluated each potential retail partner for its willingness and ability to pursue a joint goal and make any organizational adjustments that might be required to execute the plan. Teams from PepsiCo and each retailer identified data-driven ways to spur growth that would be mutually beneficial. PepsiCo teams used proofs of concept and testing to validate the effectiveness of proposed solutions. They were able to determine the effects of changes and prioritize retailer goals such as profitability when formulating solutions, even if that meant advising against more shelf space for the company’s products.
As PepsiCo’s goals became more partner-oriented, its success measures shifted from reflecting the progress of PepsiCo and its products to reflecting a triple win for the shopper, the retailer, and the company. Collectively, these practices positively influenced PepsiCo’s reputation as a partner: For the past six years, it has claimed the top spot among manufacturers in Kantar’s PoweRanking report, based on annual surveys of retail business partners.10
Manage Rollout and Value Capture
Digital solutions cannot pay off if they are not used. Organizations that succeed in strategic data sharing actively shepherd initiatives from ideation to delivery, with results in mind. They do this by drawing on innovation management practices that establish accountability for goals, create readiness for change, and ensure that solutions are deployable.
Organizations that succeed in strategic data sharing actively shepherd initiatives from ideation to delivery, with results in mind.
Elevance Health’s innovation team selected partners purposely using both an inside-out process (the team created a few priority problems and scouted for partners who could help solve them) and an outside-in process (the team established a clinical design board that reviewed incoming partnering requests). The team established technical standards for solutions so that they could be easily integrated into the corporate digital platform. At the start of each project, the team created a deployment plan that laid out an integration strategy for the project and identified the group within the company that was committed to deploying the solution. In 2021, a formal implementation team was created to work with partners and ensure that solutions became integrated into the company’s business activities and achieved their desired impact.
Schneider Electric, too, was intent on ensuring that its EcoStruxure products paid off. The company crafted a framework called the Digital Flywheel that tracks the type and amount of value that a product generated. The Digital Flywheel has four key components, representing the four parts of the EcoStruxure system: Connectable Products, Edge Control, Digital and Software, and Field Services.11 The company captures and tracks financial performance associated with each component.
How to Become Great at Strategic Data Sharing
Adopting data-sharing practices — and becoming great at them — takes time. And as we’ve shown, strategic data sharing requires some advance work on data management foundations. Our research indicates that organizations have two approaches that they can take to begin building their strategic data-sharing muscle.
The first way is to identify and pursue an appealing strategic sharing opportunity with small scope and clear intent. For example, you can try a pilot project with a trusted customer and then capture what you’ve learned before trying another. Over time, you will strengthen your organization’s ability to increase data liquidity, reduce any friction from sharing, set up win-win value propositions, and improve the likelihood of solution deployment as you adopt and learn from new sharing practices. PepsiCo worked with large retailers to codevelop solutions; when the team believed that a custom solution could be scaled cost-effectively to solve other retailers’ problems, it converted the solution into a turnkey application, often with a new user interface.
The second approach is to focus on becoming great at sharing data inside the organization, by maturing your data practices and developing a readiness to embrace future strategic data-sharing opportunities when the time is right. For example, you can revamp internal sharing to gain immediate efficiencies for employees who are heavy users of data. Fidelity Investments did this in 2019, when it initiated a four-year data re-platforming program to rationalize 100-plus data warehouses and analytics stores into a common analytics platform. Fidelity leveraged many of the same practices that enable strategic data sharing, such as preparing usable data assets and protecting them from unacceptable use. By 2021, Fidelity’s data scientists were spending more time on business problems and models than on gathering and cleaning data. The company also reduced the effort required to gather data to onboard new analytics use cases by 60% to 80% and improved its ability to tap external data.12 Fidelity first focused on internal sharing, with an internal marketplace in which local data owners contributed to strategic data assets by conforming new data to the enterprise taxonomy. Internal consumers could take data from multiple internal producers, combine the data, and build local solutions, as long as they followed core platform rules.13
So choose a route to develop your strategic data-sharing capabilities, and get started — it’s the path to maximizing long-term payback from your digital investments. You’ll be better poised to develop new business models, have more options for creating value for your customers, and be a more skillful player in digital ecosystems, whether as participant or orchestrator.
References
1. J.W. Ross, C.M. Beath, and K. Moloney, “Schneider Electric: Connectivity Inspires a Digital Transformation,” working paper 417, MIT Center for Information Systems Research, Cambridge, Massachusetts, May 2017.
2. R.W. Gregory, O. Henfridsson, E. Kaganer, et al., “The Role of Artificial Intelligence and Data Network Effects for Creating User Value,” Academy of Management Review 46, no. 3 (July 2021): 534-551; and R.W. Gregory, O. Henfridsson, E. Kaganer, et al., “Data Network Effects: Key Conditions, Shared Data, and the Data Value Duality,” Academy of Management Review 47, no. 1 (January 2022): 189-192.
3. “Sustainable Hospitality,” Schneider Electric, accessed Dec. 19, 2022, www.se.com.
4. J. Rodriguez, G. Piccoli, and B.H. Wixom, “Increase Data Liquidity by Building Digital Data Assets,” research briefing XXI-11, MIT Center for Information Systems Research, Cambridge, Massachusetts, November 2021.
5. B.H. Wixom, “PepsiCo Unlocks Granular Growth Using a Data-Driven Understanding of Shoppers,” working paper 439, MIT Center for Information Systems Research, Cambridge, Massachusetts, December 2019.
6. B.H. Wixom, G. Piccoli, I.M. Sebastian, et al., “Anthem’s Digital Data Sandbox,” working paper 451, MIT Center for Information Systems Research, Cambridge, Massachusetts, October 2021.
7. F. Lucini, “The Real Deal About Synthetic Data,” MIT Sloan Management Review 63, no. 2 (winter 2022): 11-13.
8. “Anthem Digital Data Sandbox Public Dictionary: Version 2.0,” PDF file (Chicago: Matter, n.d.), https://matter.health.
9. Wixom, “PepsiCo Unlocks Granular Growth.”
10. “PepsiCo and Walmart Claim Top Spots in Kantar PoweRanking,” Kantar, accessed Dec. 7, 2022, https://cdne.kantar.com.
11. For more information, see P. Weill and S.L. Woerner, “Dashboarding Pays Off,” research briefing XXII-1, MIT Sloan Center for Information Systems Research, Cambridge, Massachusetts, January 2022.
12. B.H. Wixom, G. Piccoli, and J. Rodriguez, “Fast-Track Data Monetization With Strategic Data Assets,” MIT Sloan Management Review, July 29, 2021, https://sloanreview.mit.edu.
13. Wixom et al., “Fast-Track Data Monetization.”