In September 2026, the Department for Business, Innovation, Science and Trade (BIST) published Smart Data Challenge Prize: Significant insights and outcomes. The programme was delivered by Challenge Works (Nesta) with the Open Data Institute (ODI) and the Smart Data Foundry.
The report brings together insights from the 10 finalists and one Smart Data Fellow who built prototypes inside a synthetic data sandbox, testing what becomes possible when people and businesses can share their data securely. The best example of this in practice is the UK’s adoption of Open Banking.
Funded by Smart Data Research UK, the Geographic Data Service (GeoDS) has been building infrastructure to address the questions the Smart Data Challenge Prize set out to answer.
A central theme of the report, and a message we consistently espouse, is that smart data is most useful when it joins up multiple sources of information — enabling better decision-making and empowering people with more choice. Across every sector that was tested by the finalists, from property and energy to transport, retail, financial services and carbon reporting, BIST concludes that “the value is not in one dataset alone”.
“Energy use can be more valuable when combined with tariff, property and weather data. Homebuying can become easier when property, identity, finance, and environmental information can be trusted and reused. Grocery services can become more personal when product information is combined with shopping history and dietary needs. Gig workers can make better decisions when earning data from different apps is brought together.”
At GeoDS, our mission is to employ geographic methodologies to integrate diverse smart data sources with nationally representative administrative and statistical data. All interactions, whether virtual or physical, happen somewhere, which makes geography the optimal medium for linking, verifying and analysing diverse sources.
Delivering smart data use cases
The report also highlights why the integration needed to deliver a smart data economy keeps stalling. Progress is constrained by inconsistent data access, limited interoperability and uncertainty around standards. “Finalists repeatedly found that promising services became harder to scale when data is held in different formats, access through different routes, or governed by different rules,” the report notes.
The Moverly use case depended on provenance attached to every claim so professionals would accept it without needing to manually re-check every input in a property pack. The Debit Score® use case found the sandbox could not supply customer-level pricing and renewal data, so live switching (such as switching mortgage providers) could not be fully replicated. The Mealia use case set aside the synthetic datasets provided within the sandbox, and fell back on using real-world data from its own retailer integrations. DIST’s own conclusion is that prototypes cannot reach scale without real data and live systems.
The report catalogues the reasons linkage fails (issues with identifiers, formats, standards and provenance), but stops short of identifying what the join in a connected smart data ecosystem actually is.
Almost every use case is anchored in place. Property packs are linked to addresses, linking EPC records to meter data is address matching, retrofit and grant targeting is focused on neighbourhoods, and multi-modal travel is movement between locations. Geography is the glue that connects all these data together.
The GeoDS team validates smart data through linkage with multiple sources and document the results, producing research-ready data, and supporting secure access through trusted research environments, while working with numerous data providers and applying FAIR principles throughout.
This is vital work, and vitally important to get right. GlowTrack notes that incomplete or outdated inputs make its risk score on fuel poverty less accurate, which matters because the index exists to prioritise vulnerable households. Coverage of smart meter, EPC and local authority records varies from place to place, and so the households that are missed in a poorly linked index are not a random sample of those in need. An integration error becomes a targeting error, which manifests unevenly in places. At GeoDS, we seek to measure such unevenness, or inequalities, across neighbourhoods, towns, cities and regions.
The report says the task ahead is turning potential into a working system. Geographic integration is the way to get this done.
Access the GeoDS Data Catalogue
GeoDS is the UK’s leading research provider of georeferenced smart data. We empower our users to generate location-driven insight and impact for projects in the public good.
Our Data Catalogue features over 100 research-ready datasets and tutorials on topics ranging from population, retail, transport, economy and digital lives.
This includes:
