Mid-market manufacturers and industrial service firms face a familiar problem: they generate valuable operational data but lack the scale to build sophisticated predictive maintenance models on their own. A growing response is the shared data trust model, where competitors and peers form industry consortiums to pool customer data for collective insight, while retaining legal ownership of their own datasets.
This explainer examines how the model works, why it is gaining traction, and what mid-market firms should consider before joining.
What is a shared data trust?
A shared data trust is a legal and governance structure that allows multiple organisations to contribute data to a common pool, managed by a neutral third party or a jointly owned entity. The trust sets rules on data access, usage, and deletion. Crucially, each member retains ownership of its own data; the trust only holds a licence to use the data for agreed purposes, such as training predictive maintenance algorithms.
The model differs from open data sharing or simple data licensing. In a trust, members have a say in governance, and the trust is accountable to them. This structure is designed to reduce the risk of one member gaining unfair advantage or of data being used beyond the agreed scope.
Why mid-market firms are turning to consortiums
Mid-market firms often have enough data to see patterns but not enough to build robust models. A single factory might have sensor data from a few hundred machines, which is insufficient to train a reliable failure-prediction algorithm. By pooling data across several firms, the combined dataset becomes statistically meaningful.
Consortiums also spread the cost of data engineering and model development. Instead of each firm hiring a data science team, they share the expense of building and maintaining the trust infrastructure. For many mid-market firms, this is the only affordable route to advanced analytics.
Another driver is the need to respond to customer demands for predictive maintenance. Large original equipment manufacturers (OEMs) and industrial customers increasingly expect suppliers to offer predictive maintenance as part of service contracts. Smaller firms that cannot meet this expectation risk losing business. A consortium allows them to offer the capability without massive internal investment.
How the model works in practice
A typical consortium might involve five to twenty non-competing firms in the same industry, such as food processing or packaging machinery. They agree on a common data schema, so that sensor readings from different machines are comparable. The trust then cleans, anonymises, and aggregates the data, removing any identifiers that could trace back to a specific customer or site.
The trust uses the pooled data to train predictive maintenance models, which are then made available to all members. Each member can apply the model to its own machines, using its own data for fine-tuning. The trust does not share raw data between members; it only shares the derived model or aggregated insights.
Governance is critical. The trust agreement specifies what data is collected, how it is used, and who can access it. It also sets out procedures for adding new members, handling disputes, and terminating the trust. Legal counsel with expertise in data protection and competition law is essential, as consortiums can raise antitrust concerns if they involve price or market sharing.
Commercial impact
For mid-market firms, the commercial benefits are tangible. Predictive maintenance reduces unplanned downtime, which is a major cost in manufacturing. A study by McKinsey (2015) estimated that predictive maintenance can reduce machine downtime by 30-50% and increase machine life by 20-40%. While these figures are often cited, they are based on specific contexts and should be treated as indicative rather than universal.
Beyond cost savings, the ability to offer predictive maintenance can be a differentiator in tenders. Customers are increasingly asking for service-level agreements that include uptime guarantees. A consortium-backed model gives mid-market firms the credibility to make such commitments.
The trust model also reduces the risk of data lock-in. Because members retain ownership, they can leave the consortium and take their data with them. This is a significant advantage over cloud-based analytics platforms where data may become trapped in a vendor's ecosystem.
Risks and unknowns
Despite the benefits, the shared data trust model carries risks. One is the challenge of data quality. If members contribute inconsistent or incomplete data, the pooled model may be unreliable. The trust must invest in data validation and standardisation, which adds cost.
Another risk is the potential for unintended data leakage. Even with anonymisation, there is a chance that aggregated data could reveal sensitive information about a member's operations or customers. The trust must implement strict access controls and regular audits.
Competition law is a further concern. In the UK and EU, data-sharing arrangements between competitors can be scrutinised under antitrust rules. The trust must ensure that it does not facilitate the exchange of commercially sensitive information such as pricing or customer lists. Legal advice is essential.
Finally, there is the question of model performance. A model trained on pooled data from multiple sites may not perform well on a specific member's machines if their operating conditions differ significantly. Members may need to invest in local fine-tuning, which reduces the cost savings.
Why it matters
The shared data trust model is part of a broader shift towards collaborative data governance. As data becomes a critical business asset, firms are looking for ways to share insights without losing control. The trust model offers a middle path between keeping data siloed and handing it to a large platform.
For mid-market firms, the model could level the playing field with larger competitors who have in-house data science teams. It also aligns with regulatory trends that favour data portability and user control, such as the UK's post-Brexit data regime and the EU's Data Act.
FY Outlook
Expect to see more industry-specific consortiums emerge over the next two to three years, particularly in sectors with high-value machinery and strong safety requirements, such as manufacturing, energy, and logistics. The success of early adopters will determine the pace of adoption.
We also anticipate the development of standardised trust frameworks and third-party trust managers, which will reduce the legal and technical barriers to entry. However, the model will not suit every firm. Companies with highly proprietary processes or those in highly competitive markets may find the risks outweigh the benefits.
Conclusion
The shared data trust model offers mid-market firms a pragmatic route to predictive maintenance without surrendering data ownership. It requires careful governance, legal oversight, and a commitment to data quality. For firms that can navigate these challenges, the potential commercial upside is significant. As the model matures, it may become a standard tool in the industrial data toolkit.
Source notes
Editorial note: The McKinsey figure on downtime reduction is widely cited but originates from a 2015 article. We have not verified the underlying data and recommend treating it as indicative.
Editorial note: For current examples of industry consortiums, we recommend reviewing public case studies from the UK's Digital Catapult and the EU's Data Spaces initiative. No specific URLs are provided here to avoid unverified links.
Editorial note: Legal considerations around data sharing and competition law are based on general principles. Specific advice should be sought from qualified counsel.
