Sharing goes mainstream, and the news is full of data sharing agreements between companies and within industries. In addition, intermediaries are every year more numerous to position as enablers of data sharing. When taking a closer look at these initiatives, one immediately concludes that not all sharing agreements are equal and strategic motives govern each. It’s not about « to share or not to share » but rather « what to share and what not to share ». How can data sharing be a strategic tool to change the balance of power in an ecosystem?
Sharing goes mainstream
There is an increasing interest in data sharing, at least because companies see that deriving value from data implies gaining access to multiple data sources and being able to mix and analyse various datasets. A recent study commissioned by the European Commission stated that, in the EU27 member states, the flow of data stemming from enterprises buying cloud services over the internet was 5.8 exabytes a year in 2020. By 2025, forecast to be almost five times greater and even 15 times greater by 2030.
The building blocks and key enablers are here: the emergence of open standards (facilitating interoperability), the multiplication of data platforms (such as Dawex) and new data intermediaries, a legal framework that aims to foster trustworthy data sharing at the European level (the European Data strategy, 2020), the Gaia-X initiative: all these contribute to overcoming the fears of private actors.
Sharing takes multiple forms, and we focus our analysis on voluntary-based and strategically-motivated data sharing initiatives. That means we excluded open data as it implies no direct compensation, pure monetary-based sharing (such as data marketplaces) and law-enforced data sharing (such as for the exchange of passenger name records – PNR – in the airline industry).
Not all sharing initiatives are created equal
In this paper, we see data not as a product to be sold but as a strategic asset of organisations. As a resource for a company, there are both incentives to keep it as exclusive as possible and incentives to share it with other stakeholders and competitors. Let’s have a more detailed look at this alternative. Exclusivity has always been a source of competitive advantage. Suppose a company is the only one accessing a specific resource (be it a material such as rare metals or exclusive knowledge and capabilities). In that case, it’s in a better situation than competitors. One should mention that even when data is shared, there’s still a part of exclusivity. For instance, one could choose to share historical data and to keep access to real-time data exclusive. Therefore the question is not really « to share or not to share », but how far do you share and under which configuration?
Let’s review some examples to illustrate our argument:
- Amazon marketplace shares individual merchant information with each merchant; however, it doesn’t share benchmark information (all merchants of the same category) nor the algorithm used for ranking the products. In this case, the sharing configuration 1) helps merchants to grow their business and 2) preserves Amazon’s interests through asymmetry of information.
- With Walmart retail link, the retailer shares with its suppliers point-of-purchase cash register data so that they decide on delivering new products according to accurate and precise consumption information. This detailed data is precious to brands selling at Walmart; they know better when and where their products are purchased, and Walmart trades that information against the cost to keep the inventory which is transferred to the suppliers.
- Waze connected cities’ program aims to establish a bi-directional data exchange between Waze and the municipalities. To get access to traffic data collected by Waze, local public authorities need to share data about road closures due to planned events or road maintenance work. In this case, the sharing configuration 1) helps cities access traffic data without exchanging money and 2) benefits Waze, which gets premium access to information about the road infrastructure, which in turn could be used to increase its own product/app and differentiate against competitors.
- Driver’s seat is a rideshare analytics cooperative that drivers own. In exchange for data about their activity (automated tracking), drivers get insights into their performance. Driver’s seat adopts a shared ownership model and derives revenues from selling gig data to organisations with similar drivers-first missions. In this case, the sharing configuration 1) helps drivers get insights that they would never be able to get alone and 2) it reduces the asymmetries of information where only platforms (such as Uber) have a global view of the economics of the gig economy.
- Catena-X aims to « provide an environment for the creation, operation and collaborative use of end-to-end data chains along the entire automotive value chain ». Originating in Germany, Catena-X gathers the leading German carmakers and IT specialists and software editors to develop a vertical data space. The goal is to go beyond the traditional electronic data interchange (EDI) already well implemented in the automotive supply chain. In this case, the sharing configuration 1) reduces the cost of data exchange by introducing interoperable data standards and 2) is supported by dedicated multi-stakeholders governance.
- Credit bureaus, such as Schufa in Germany, aim to reduce the risk for lenders by sharing credit data on loan applicants. Each banking and credit industry company can share individual information about their customers. In return, Schufa will then calculate an individual score for each individual or household based on its credit history. In this case, the sharing configuration is based on 1) a centralised organisation (credit bureau) that gathers individual information and 2) a scoring system that turns this information into an actionable insight that companies can use for many applications.
5 Dimensions to define sharing configurations
The above examples lead us to identify five dimensions to define a sharing configuration. The first three dimensions refer to the asset shared, and the other two refer to both parties’ incentives.
Three dimensions on the asset
Sharing configurations are defined by choices made on the asset:
- What granularity and scope of data are shared? Data can be exchanged raw, cleaned, aggregated or calculated. The more granular the data shared, the more flexibility is offered to the data consumer and the more effort is required to transform the data into insights. This flexibility makes the data more attractive, which triggers more adoption, but at the same time, it reduces the exclusivity of the source for the owner. With more granular data, the data owner trades adoption against control. In the Walmart retail link example, very granular data is shared with suppliers in exchange for the transfer of the inventory-keeping activity.
- Who can access data? The Open Data Institute synthesised the options in the open data spectrum. Data can be shared with identified parties, within a group or publicly. The more public the data, the less exclusive, and the decision of which parties could access it defines competition and cooperation spaces. In the example of Driver’s seat, the sharing takes place among the drivers’ members of the cooperative and is extended to the organisation which shares the drivers-first mission. Data sharing is a way to build and sustain collaborations within an ecosystem. In the Connected Cities example, the data is accessible only to cities that have signed a specific agreement.
- What uses are permitted for the data? When data is shared, a licence is granted to the consumer. Licences can give several rights: the right to use, modify, distribute and monetise. The more rights granted, the more attractive data are for the consumer. And more value is transferred from the owner to the consumer. In all the examples of our sample, rights are clearly defined and somehow limited.
Two dimensions on the incentives
In addition to the dimensions on the asset and for a sharing initiative to play a strategic role, there are two considerations to explore which will influence the choice of the options. The first has to do with the organisation that shares, and the second with the organisation that uses the shared data.
- What benefits for the data owner? The organisation that decides to share data benefits from clearly identifying its expectations towards this sharing strategy to deliver tangible strategic results. It could be to use the data shared as a currency to access another resource. Waze benefits from a very valuable resource from cities in exchange for Waze users’ aggregate data. Sharing data with suppliers and helping them better manage their operations puts Walmart in a better position towards its suppliers than its direct competitors.
- What benefits for the data user? We know from the sometimes disappointing results of open data portals that it’s not only because some asset is available that organisations rush to use it. This means that for a company to leverage data sharing as a strategic tool, it must clearly investigate and define the benefits of using the shared asset. Companies using the data gathered through Catena X may see their operations improved with new insights at other parts of the value chain. Sellers on Amazon may use the data shared to enhance their marketing actions. Banks using credit bureau data would better assess risks and reduce their costs.
Thinking strategically about sharing
The two groups of dimensions, on the incentive and on the asset, interact. According to the incentives anticipated on both sides (data owner and consumer), different configurations may be adopted on the asset. For example, when data is used as a currency to access another data resource, the data shared is likely to be more granular, the scope more extensive, the consumers limited and identified and the rights excluding monetisation and distribution.
If sharing is a strategic tool, then here are some helpful questions to keep in mind when defining a sharing configuration:
- How could sharing data with competitors improve the bargaining power against suppliers or clients?
- How could sharing data with suppliers or clients help differentiate against competition and capture market share?
- Under which configuration would sharing data offer high cost or risk reductions?
- What other resource or activity could be traded in exchange for data?
- How can partners negotiate sharing configurations to strike a more favourable balance?
Written by Louis-David Benyayer and Simon Chignard