Data-sharing rights, delivery costs and commercial dataset value are different variables. This Finndy Research worksheet examines how to separate them for an enterprise considering connected-product data access in Europe. The European Commission’s Data Act overview confirms application from 12 September 2025. Its Data Governance Act explanation describes a distinct framework for trusted voluntary sharing. The numbers below are transparent hypothetical assumptions, not quotes from sellers or legal interpretations for a particular company.
Start with the legal entitlement, not the asking price
A supplier cannot price a data contract responsibly until it separates voluntary commercial licensing from sharing that a law may require. The European Commission explains that the EU Data Act entered into application on 12 September 2025 and addresses users’ rights concerning data generated by connected products and related services. That is not a blanket rule giving every buyer every dataset for free, and it is not permission to resell all personal or confidential information. The analyst should first map the holder, product user, eligible recipient, geographic scope, dataset, purpose and applicable contractual restrictions. Create a decision record with three outcomes: mandatory-access candidate, voluntary-license candidate or legal review required. Keep these branches visible because a voluntary dataset market valuation and a cost-based assessment of making data available answer different questions. If parties disagree about eligibility, postpone the final price rather than quietly treating an asserted legal right as a conventional premium product.
Construct an auditable delivery-cost baseline
For a hypothetical industrial telemetry feed, write down the costs directly caused by preparing and delivering one contract rather than applying an unexplained revenue multiple. A practical ledger contains extraction and connector setup, documented transformations, access controls, security review, incremental bandwidth, delivery monitoring, customer support and independently verifiable quality checks. Allocate each recurring cost by a stated driver: records delivered, gigabytes transferred, refresh cycles, authorised recipients or support hours. For example, assume an illustrative month with 200 euros of connector maintenance, 120 euros of security operations, 80 euros of delivery costs and 100 euros of review work. The total is 500 euros before any assessment of legally permitted compensation or negotiated margin. These are constructed assumptions, not market prices. Record excluded shared overhead, sunk investment and already-funded collection costs separately, so the same expense is not recovered twice under different labels.
Keep compensation logic distinct from resale value
The Council of the European Union described reasonable compensation and unfair-contract safeguards when adopting the Data Act, while the Commission’s current explanatory material points to future guidance on reasonable compensation for mandatory business-to-business sharing. The mere existence of a high downstream use-case value therefore does not establish a lawful mandatory-sharing fee. Use two worksheets. Sheet A records legal access category, eligible cost components, factual assumptions and specialist review. Sheet B models voluntary commercial licence value using buyer alternatives, timeliness, documented coverage, expected decision improvement and contractual restrictions. Do not move a price from Sheet B into Sheet A simply because a buyer claims the data will improve margins. Commercial value can inform negotiation where negotiation is lawful, but it must not replace the governing legal analysis. This separation also protects ordinary pricing research from being misrepresented as legal advice.
Quantify quality and refresh without pretending precision
A dataset can be commercially useful and still unreliable for a particular decision. Define acceptance tests before forecasting revenue: coverage by device, missing-value rate, timestamp delay, schema stability, provenance, deletion handling and correction latency. Consider two stylised offers delivering the same nominal million records. Feed Alpha has 98 percent usable records and a one-hour refresh; Feed Beta has 90 percent usable records and a 24-hour refresh. These figures are examples, not observed supplier performance. Rather than proclaiming Alpha is worth 8 percent more, estimate the buyer’s incremental validation burden, rejected-record handling and operational delays. A model might calculate adjusted usable units as delivered units multiplied by the tested acceptance fraction, then add documented cost of freshness. Report sensitivity at several thresholds. This method makes disagreement about usefulness testable instead of translating every quality attribute into an arbitrary premium.
Design contract boundaries around revocation and rights
The European Commission’s explanation of the Data Governance Act emphasises trust in voluntary sharing arrangements and the particular obligations of data intermediation services. That framework is not interchangeable with the Data Act. For any cross-enterprise data service, prepare a rights matrix listing source authority, allowed field groups, permitted purposes, recipients, retention, derivatives, audit logs and revocation conditions. Build a technical evidence trail of consent or lawful basis, authentication, delivery receipt and deletion requests. Contracts should state whether cached derived results survive termination and how disputes are reconciled. Do not infer that holding a database grants unlimited intellectual-property or privacy rights. A data marketplace may handle transport and administration while leaving legal permissions with underlying parties. Each product needs a documented route for suspending access when the rights record changes.
Run scenario and dispute tests before quoting
The final pricing model should survive disagreement, not just a favourable forecast. Prepare at least three scenarios: baseline delivery, a doubled extraction burden and a failed quality acceptance month. Use an illustrative cost-to-serve baseline of 500 euros, a stress case of 720 euros and a remediation case with an additional 160 euros of review. Do not automatically label these amounts permitted statutory compensation; legal constraints and contract terms must be assessed independently. For a voluntary licence, compare the price to buyer alternatives and model how delayed updates reduce realised utility. Show every input, currency, tax treatment and review owner. Red-team the spreadsheet with disputes over excluded devices, partial deletion, duplicate invoices and a source-data correction after resale. A credible quote tells both parties what was measured, what is assumed and which claims still require legal or technical verification. That transparency is itself part of the data product’s value.
Decision rule: document entitlement first, verify delivery economics second, and model voluntary licence value only inside the rights that actually exist. Revisit assumptions when regulation, infrastructure, or buyer requirements change.
Sources (reviewed 13 November 2026): European Commission, Data Act (application date 12 September 2025); European Commission, Data Act explained (current implementation guidance); European Commission, Data Governance Act explained (applies since September 2023); Council of the EU, adoption note, 27 November 2023.
