We don’t even think about it. You pull into a gas station and insert your credit card. That was a major cross-industry innovation. Essentially putting an ATM machine onto a gas pump enabled the pay-at-the-pump, self-service gas station. That first combination appeared at a gas station in Abilene, Texas, in 1973. Interestingly, by 1994, only about 13% of convenience stores offered pay-at-the-pump. Yet less than a decade later in 2002, the percentage of convenience stores with pay-at-the-pump jumped to 80%. According to the Petroleum Service Company, it took collaboration with another industry, telecommunications, to truly scale the innovation.1
Cross-industry collaboration historically has led to many transformational breakthroughs we all recognize and have benefitted from. A big one is the fusion of glass, cable, and electronics into fiber optics, which accelerated access to high-speed internet. Now cross-industry data collaboration is becoming increasingly common as well. Car manufacturers share telematics data with insurance companies to optimize and personalize premium pricing. Retailers share data with consumer packaged goods (CPG) producers to improve product development and demand forecasting. Telecom operators work with financial services organizations to improve credit scoring and deliver mobile banking services.
Necessity often breeds innovation. The pandemic extended the boundaries and accelerated data access and collaboration across industries and geographies. The World Health Organization’s “COVID-19 Open” data sharing and reporting protocol established a framework for data sharing during a crisis.2 According to Reuters, telecom operators shared data with governments looking to track mobility and potential contagion.3 Tour operators shared data on flight cancellations and country-specific restrictions to improve their ability to update customers about rapidly changing travel environments. This was an exceptional time, but some of those data sharing behaviors will remain.
As we return to a post-pandemic world, it will no longer be enough to break down internal data silos to share across business units. Companies will increasingly look further afield for opportunities to blur industry boundaries to solve business challenges and deliver value. We expect more cross-industry collaboration.
The New Frontier: Secure Data Collaboration
The pandemic has cast a spotlight on the healthcare and life sciences industries, both beacons of cross-industry collaboration. Pharma’s drug development is at the crossroads of healthcare, itself a multi-industry endeavor, and manufacturing. Driving innovation requires reconciling complex data relationships between disparate, siloed data sources such as de-identified patient records, clinical data pools, physician registries, pharmacy records, and more. The rapid development and delivery of COVID-19 vaccines demonstrate the benefits of data sharing. But in healthcare, data sharing is not always straightforward.
The promise of personalized medicine requires data collaboration. Across the patient lifecycle, data must be shared to predict individual responses to disease and treatment. Hospitals, clinics, and pharmacies share data to get a complete picture from diagnosis to treatment to outcome: the patient 360. Yet, personalized medicine starts with personal data, which often cannot be shared. New technologies enable hospitals to share their data without compromising the privacy of individuals. These methods are referred to as secure data collaboration.
In the financial services industry, insurance companies and banks want to detect fraud and can benefit from cross-industry data sharing to better understand the profile of fraudsters and fraudulent transaction patterns. A large U.S. insurance company shares its fraud models with other insurers to enable it to train its machine learning models on more data. Illustrating a more cross-industry example, pharmacies and insurance companies share data to identify claims and reimbursement fraud. Secure data collaboration allows participants to benefit from data sharing without violating privacy or industry regulations.
In the energy and utilities industry, data collaboration isn’t new but opportunities are expanding. Energy performance contracts are often based on cooperation between an insurance company, a bank, a construction firm, and an energy equipment installer. When renovating or retrofitting an older building, projects are financed through future cost savings. The equipment provider and construction firms guarantee performance, banks finance the project, and the owner repays the loan based on the cost savings over time. Data sharing enables these cross-industry risk assessments to support this kind of conditional business financing.
Now, smart meters in individual homes literally open new doors for data collaboration, transmitting information about energy consumption. Analysis of the data can help consumers compare themselves to others and change their consumption behavior. Utility companies can better predict demand and ensure adequate supply. Because the data is associated with individuals, secure data collaboration is required: sharing the data without allowing others to see it.
Seeding the Cloud with Data to Accelerate Innovation
The Snowflake Data Cloud enables companies to think outside their industries and identify new opportunities for innovation through data sharing. Retailers and CPG manufacturers can share data to improve product development, better forecast demand, and optimize inventories and prices. Advertisers and media outlets can share data to personalize advertisements and increase impact. In healthcare, connecting patients, doctors, pharmacies, and insurers improves health outcomes and lowers costs.
In these examples, however, there are also potential data security risks and privacy concerns. But there are ways companies can share information without showing their data. The idea was first launched in the early 80s when a professor Andrew Yao posed the question: How can two people share information without showing each other their data?4 In the scenario posited, two millionaires wanted to know which one had more money. But neither wanted the other to know exactly how much they had. They wanted to share but not show. Sounds impossible, right?
Wrong. It is possible. Mathematicians figured it out with a very complex proof. Fortunately, computers can do it better. Snowflake provides the tools that allow companies to encrypt their data and collaborate using double-blind joins and secure user-defined functions. Users can get the results when they decrypt their data, while never actually seeing the other’s data. The two millionaires know which is wealthier without knowing by how much. In the scenarios discussed above, hospitals, insurers, or utility companies can pool encrypted data and perform analytics to help personalize medicine, detect fraud, or forecast energy demand.
Stay tuned for more from Snowflake about secure data collaboration, or data clean rooms (although they’re not rooms at all).