B2B data sharing: Driving Business Insight and Innovation

Data sharing between businesses has always been an issue. The primary reason for this is the lack of trust between organizations. Businesses like to ensure that their markets are not negatively affected by sharing sensitive corporate data, which acts as a set-back to data sharing. Data ownership is nowadays perceived as equally important to owning oil or physical assets. Therefore, many businesses consider the confidentiality of their datasets as a critical factor that affects their ability to maximize their market share through their current data-driven business practices. In this context, individual businesses seek benefits from data sharing only when they can do so without letting their competitors on the game. Furthermore, there are also cases where companies are reluctant to share data with their business partners because they afraid that this could put their customers’ privacy and security at risk.

These are certainly valid concerns when companies share data on public portals or deploy poor cybersecurity measures. Nevertheless, it is no longer true that data sharing cannot be safe.  Recent advances in data-sharing infrastructures and privacy preserving technologies (e.g., privacy preservation computation) make it possible for businesses to securely share data with their business partners in various supply chains where they participate. Such technologies are already used in a variety of use cases that are based on the sharing of sensitive business data, such as use cases in the areas of digital finance, healthcare, and industry.

Prominent B2B Data Sharing Technologies

The emergence of new data sharing technologies is gradually changing the way companies think about data sharing in business-to-business (B2B) settings. Here are some of the emerging technologies that make the difference:

• Data Spaces: Data spaces are federated decentralized databases of different organizations. They are data sharing platforms that enable authorized individuals to securely access data from other organizations. In more practical terms a Data Space is a decentralized database that is shared across multiple organizations. It allows authorized users from each organization to share information with each other without having access to the same physical resources on their own servers. This means that users can securely access any piece of information in the database without having direct access to any other user’s system. This makes it possible for organizations that do not share infrastructure (e.g., cloud data centers) to share information over a network without losing control over their own assets.
• Data Marketplaces: Data marketplaces include repositories of datasets from different organizations. They enable the trading of different datasets across various market actors including for example customers, suppliers and other business partners. Data marketplaces are like online shopping malls where you can find all sorts of products and services. They are platforms where data providers can sell their data to interested consumers (i.e., users) of these data. Note that data marketplaces go beyond simple B2B data catalogs as they offer more trading capabilities over a wide range of data-driven products and services such as raw data, data analytics insights, machine learning models and many more. One of the main goals of data marketplaces is to allow businesses to share data assets in a secure and efficient way. They also offer transparency over the available data assets, which encourage companies to share even more data in order to benefit from new insights and innovations. Such innovations would be hardly possible without sharing datasets with other companies.
• Data Interoperability technologies: Data interoperability technologies are key to combining data assets from multiple sources. Nowadays, companies are increasingly working with an increasing number of data sources. This is due to the rise in digital transformation, which has led to the proliferation of new digital platforms. These platforms have enabled companies to collect a variety of datasets from different sources, including Internet of Things (IoT) sensors, social media streams, e-commerce systems and more. In order for businesses to maximize the value of their data, they need to be able to combine information from the various sources and to make the combined datasets available across their organization. This is where data interoperability comes into play. Data interoperability technologies enable the combination of different datasets from a variety of sources despite their different semantics and formats. They are used to make data accessible across multiple applications, unify disparate data sources into a single view, and help ensuring that data is accurate and up-to-date. Furthermore, data interoperability technologies facilitate data-driven collaboration across departments of an organization, notably departments that possess different datasets.

Privacy Preserving Computation Techniques

Except for these technologies, there are also advances in tools and techniques for preserving the privacy of customers and other end-users. Specifically, modern organizations can use privacy preserving computation techniques to overcome privacy and data protection challenges. These techniques help companies to analyze data in a way that ensures confidentiality while still driving business insights and innovation. They come to alleviate the ever important inherent conflict between data sharing and privacy protection, which stems from the fact that companies have no control over who has access to their data.

Privacy Preserving Computational Techniques enable privacy friendly analysis of different datasets in ways that do not expose the source data, but only the results of the analytics and the queries. They do this by leveraging cryptographic protocols for secure computation, which allow computation on encrypted inputs without revealing either the inputs or the interim results of computations on them. This enables privacy friendly analysis of different datasets in ways that do not expose the source data.

Privacy preserving computation approaches enable companies to share data with their partners in order to gain insights into their own business, or access to new opportunities. From a technological perspective, privacy preserving computation uses a cryptographic technique that is called ‘homomorphic encryption’. The latter allows computations on encrypted data without revealing anything about the underlying plaintext data. This makes it possible for two parties to collaborate on a project without ever sharing any sensitive information.

The execution of homographic encryption computations is very demanding in terms of computational resources. Once upon a time, enterprises were not offered with cost-effective access the compute cycles needed to implement non-trivial privacy preserving computations. This has recently changed with the broad availability of large amounts of compute cycles at an affordance cost.

While there are multiple challenges facing B2B data sharing and analyses, there is also an increased focus on developing solutions for these challenges. Organizations that implement solutions for B2B data sharing initiatives will be able to improve their business processes and gain competitive advantages over others which do not invest in this area. Specifically, companies must bear in mind that open sharing of data is a key to harnessing the true potential of big data in the data economy. This is the reason why enterprises must seriously consider establishing a successful data sharing infrastructure that will help them share data with their business partners in effective ways and with guaranteed data security.