Data Integration & Ontologies

How do you choose a data platform for a property portfolio?

Test platforms on your hardest buildings and on exit, not on the demonstration dashboard.

Conceptual isometric illustration: Comparing platforms across hotel and office pilots-2, not a verified engineering diagram.
On this page
  1. Describe the portfolio you actually have
  2. Score on evidence, not on promises
  3. Test the exit while you still have leverage
  4. Pilot on your most awkward building
  5. Someone has to own it afterwards
  6. Practical application
  7. Sources

A supplier's demonstration naturally shows its platform at its best. Your portfolio will be less obliging. A portfolio data platform should be chosen by testing it against the information and workflows you need to keep for years, across a representative mix of your own buildings. Integration effort, data governance, operating cost and the cost of leaving are what usually decide whether a platform still pays its way three years after rollout.

"Data platform" here means the service that collects building data from many sites, gives it consistent identities and context, stores the history and makes it available to applications and people. It may be one product or several.

Describe the portfolio you actually have

Requirements should start from a representative sample of buildings: new and old, different BMS makes, good and poor documentation, managed and multi-let. For each one, list the systems, the data needed, the history required and who supports it.

From that sample, the requirements can set out the core information the owner must retain, meaning identities, relationships, history and access; which applications must work from day one and which can follow; the interfaces expected, including documented APIs (for example, described with the OpenAPI Specification), export formats and the semantic model; what must keep working on site if the connection fails; and the governance arrangements for roles, permissions, personal data and audit.

Essentials and attractive extras should be kept apart. A feature nobody is responsible for using should not carry the same weight as reliable access to core information.

Score on evidence, not on promises

A weighted scoring sheet keeps the evaluation honest. The weights below are a starting point only; set your own before the bids arrive.

CriterionExample weightEvidence required
Integration of the pilot buildings25%Working connections in the pilot, effort recorded
Data model and quality controls20%Delivered model, validation results, quality flags
Openness and exit20%Documented API, full export tested by a third party
Whole-life cost (5 years)20%Licences, connections, storage, support, internal staff
Security and governance10%Access model, audit, hosting, supplier assurance
Usability for intended users5%Tasks completed by your own staff in the pilot

Cost needs probing beyond the headline. Ask how charges change as buildings, users, data volumes and third-party applications grow, because per-point, per-API-call or per-application fees can change the picture substantially at portfolio scale.

Test the exit while you still have leverage

During the pilot, require a full export of data, relationships and configuration, and have someone other than the supplier load it into another tool. The EU Data Act (Regulation (EU) 2023/2854), which has applied since 12 September 2025, includes rules to help customers switch between data processing services, including cloud services, and in November 2025 the Commission proposed targeted changes to those rules as part of its Digital Omnibus package. Contract terms should still spell out formats, timescales and assistance, and your legal advisers should confirm which text applies and how.

Pilot on your most awkward building

The choice of pilot sites can decide the outcome on its own.

Someone has to own it afterwards

Finally, decide who maintains connections, curates the data model, resolves data defects, manages users and handles supplier changes. A platform without an owner degrades quickly, whatever its features.

Practical application

Build the scoring sheet before meeting suppliers, and give the pilot your most awkward building, not your best; a tidy office flatters every platform. Pagin Consulting runs that process with owners as an independent, Vendor Agnostic adviser.

Sources

This article connects to our work on Our services.

Related articles

Data Integration & Ontologies

How can building owners avoid vendor lock-in?

Where vendor lock-in actually sits in building systems, how to make exit measurable and how to price access over the whole life.

4 min read · Published 24 March 2026

Related service

Our services

We can discuss how this subject fits the project brief.

Contact Pagin Consulting