The CFO's Case for Open Data Infrastructure

Your data platform is a financing decision disguised as a technology roadmap. If switching costs are unmeasurable, a vendor can quietly set your future margin.

Why finance should care

CFOs usually see the invoice. The bigger bill shows up later as migration labor, duplicated tooling, project delays, and risk premiums on delivery dates. It also shows up as lost option value. You cannot pursue the cheaper engine, the better catalog, the new AI workflow, or the second cloud because the architecture makes change slow and expensive.

Open data infrastructure (ODI) is the set of decisions that keep those costs legible. It is not a preference for open source. It is an insistence that the durable data contracts remain portable, inspectable, and governable even when tools change.

A CFO-grade cost model

If you want a CFO-grade model, stop organizing the math around vendor SKU names. Organize it around cost drivers your organization can control.

Storage: object storage, request charges, retention, compaction, and recovery behavior.
Compute: query engines, transformations, streaming, and AI workloads.
Data movement: ingestion, replication, cross-region traffic, and outbound data transfer charges.
Control plane: catalog operations, metadata management, governance enforcement, and lineage.
Operating drag: custom connectors, platform-specific SQL, and workflows that only work inside one vendor boundary.

ODI changes the model because it moves the durable contracts out of product-specific systems and into shared interfaces: open table metadata, catalog APIs, and portable governance signals. You still pay for infrastructure, but you pay less for dependency.

Core idea: the finance win is not cheaper technology. It is cheaper change.

Lock-in as financial risk

Lock-in becomes a finance problem when it turns into unpredictable cash needs. You do not budget for it the way you budget for normal platform evolution. It arrives as an emergency migration, a sudden re-platforming program, or a quarter where every critical delivery depends on the same vendor escalation path.

From a CFO perspective, lock-in risk looks like:

Unpriced exit cost: the organization cannot estimate the cost to leave until it is already urgent.
Weak negotiation power: price changes cannot be arbitraged because the switching timeline is too long.
New billing dimensions: a new workload (especially AI) introduces costs that the old model did not capture.
Compliance rework: controls, audit trails, and lineage become expensive to rebuild after data moves.

ODI is how you keep an exit path credible. Not a slide deck. A tested, operational procedure with real governance coverage.

The controls that change the math

The simplest way to explain ODI to finance is to treat it as boundary ownership. You want the organization to own the parts of the contract that are expensive to recreate.

Open table formats for durable data: keep table semantics close to the data, not trapped in a proprietary storage layer.
Open catalog boundaries: avoid private coordination paths for discovery, access, and governance decisions.
Portable metadata and lineage: treat trust signals as part of the asset, not as an internal tool detail.
Compute optionality: design so you can add a second engine without rewriting the core contract layer.

Finance does not need to own these decisions. Finance needs to sponsor the incentives that keep teams honest, including the budget for operating the boring parts: audits, backups, policy enforcement, and periodic exit drills.

Questions to ask

Which parts of our data platform spend are paying for runtime, and which parts are paying for dependency?
Can we export the data and the metadata, policies, and lineage that make it trustworthy?
What is the expected cost of switching engines or vendors for our top three data domains?
What new costs show up when we add AI agents and operational workflows?
Do we have a tested exit path, or a theoretical one?

If you want a CFO definition of ODI, use this: it is the architecture that keeps future negotiation power on your balance sheet.

Sources to start with

These sources are useful starting points for the cost and portability claims in this piece.

ODI hub Article library Use the scorecard Lock-in cost model TCO model