Voluntary carbon registries are the market's ledger systems. They validate methodology compliance, authorize credit issuance, assign serial numbers, track ownership transfers, and record retirement events. For corporate buyers and sustainability auditors, the registry record is the primary document of authority — the closest thing the voluntary carbon market has to a title deed. Understanding how that record is constructed, and where the data trail goes thin, is essential for anyone responsible for verifying the carbon claims in a corporate sustainability report.
We've analyzed the publicly documented issuance workflows of the four major VCM registries — Verra (VCS), Gold Standard, the American Carbon Registry (ACR), and the Climate Action Reserve (CAR) — focusing specifically on what data is captured at each stage of the issuance process and what is either absent or inaccessible to downstream parties. The gaps are not idiosyncratic to any single registry. They're structural features of how these systems were designed, and most of them predate the level of institutional scrutiny the market now faces.
The Issuance Workflow: What Each Stage Produces
All four major registries follow a broadly similar workflow for project registration and credit issuance, derived from their respective standard frameworks. The stages are:
- Project listing — the project developer submits a Project Description (PD) or Project Design Document (PDD), identifying the project boundaries, the applicable methodology, the additionality argument, and the baseline scenario. The registry's technical team reviews for methodology compliance.
- Validation — an accredited VVB reviews the PDD and confirms that the project's design is compliant with the methodology and eligible for registration. The VVB's validation report is submitted to the registry.
- Registration — the registry formally registers the project, assigns a project ID, and opens an account in the registry's tracking system. From this point, the project can enter monitoring periods.
- Monitoring and verification — the project developer conducts monitoring per the approved monitoring plan and produces a monitoring report. An accredited VVB reviews the monitoring report and issues a verification statement confirming the net verified quantity.
- Credit issuance — the project developer requests credit issuance for the verified quantity. The registry reviews the issuance request for administrative completeness and issues credits (as VCUs for Verra, Gold Standard Certified Emission Reductions for Gold Standard, Emission Reduction Tons for ACR, Climate Reserve Tonnes for CAR).
- Transfer and retirement — credits are transferred to buyer accounts and eventually retired, at which point they are permanently cancelled from the registry.
This framework is well-designed for its original purpose: managing methodology compliance and preventing double-counting within a single registry's project universe. What it doesn't do is capture the measurement-origin data from step one of the actual carbon quantification process, and what it doesn't enable is cross-registry comparison of the underlying physical claims.
Where the Data Trail Goes Thin
The Measurement Origin Is Not Registry-Captured
The most significant gap in all four registries' issuance records is the absence of standardized, machine-readable documentation of the measurement inputs that underlie the verified quantity. A registry issuance record contains the project ID, the vintage year, the verified quantity, the VVB name, and the monitoring report reference. It does not contain a structured reference to the remote sensing data products (SAR acquisitions, optical imagery, LiDAR point clouds) that those monitoring reports are based on.
The underlying measurement data exists — it's typically in appendices to the project developer's monitoring report or in the remote sensing firm's data delivery documentation. But the registry's database doesn't index it, link to it, or validate its consistency with the claimed quantity. The relationship between "VCU number 4,200,001 through 4,380,000 were issued on [date]" and "the Sentinel-1 SAR composite from [acquisition dates] for polygon [coordinates] showed AGB density of [X] Mg/ha" exists only in PDF documents, not in the registry's structured data.
VVB Report Data Is Not Normalized
All four registries require that verification reports be submitted as part of the issuance process, and those reports are (with some delay) publicly available through the registries' project databases. However, the content and format of verification reports varies significantly between VVBs and between monitoring periods for the same project. There is no standardized schema for a carbon project verification report that would allow the key data elements — monitoring period, verified quantity, material findings, data review scope — to be extracted programmatically and compared across projects or vintages.
This means that an analyst trying to compare the verification rigor of two projects from different registries must read both reports as narrative documents and manually extract comparable information. For a large corporate buyer with a portfolio of 50+ retired project vintages, this is prohibitively labor-intensive. For an audit firm tasked with spot-checking a client's carbon procurement, it means the verification work is done on a sampling basis rather than a complete-portfolio basis, which creates residual risk.
Beneficial Ownership at Retirement Is Weakly Documented
Registry retirement records identify the account holder who retired the credits. For credits that move through brokers or aggregators before reaching a corporate buyer, the registry's retirement record names the entity that held the account at time of retirement — which may be the broker's account, not the ultimate beneficiary corporate buyer. The GHG Protocol's guidance on retirement documentation states that the retiring entity should be the entity making the climate claim, but the mechanics of how "beneficial owner" retirement is documented varies by registry.
Verra's registry allows the retirement record to include a "beneficiary" field separate from the account holder, which is a useful mechanism. Not all registries have equivalent functionality, and even where it exists, the consistency with which it's used varies by transaction type. For a corporate buyer who purchased through a broker, confirming that the retirement record properly attributes the retirement to them (not to an intermediary's account) requires manually reviewing the retirement certificate rather than querying a structured database field.
The Structural Reason for These Gaps
It's tempting to interpret these data gaps as negligence or insufficient attention from the registry operators. That reading is incorrect. The gaps are structural features of how voluntary carbon registries were designed in the early 2000s, when their primary function was preventing double-issuance within their own project universe, not enabling cross-market provenance verification.
Verra launched the VCS Program in 2005 at a time when the voluntary carbon market was predominantly served by project developers selling directly to individual corporate buyers through bilateral agreements. The registry infrastructure was built for that context: manage project compliance, issue serial numbers, prevent the same tonnes from being issued twice within the registry. The need to expose measurement-origin data in machine-readable formats to arbitrary third-party verifiers wasn't in the design brief.
That context has changed dramatically. The market is now institutional. Corporate buyers have compliance obligations under frameworks like TCFD and the SBTi that require them to document and disclose the quality of their carbon procurement. Audit firms are reviewing carbon retirement portfolios as part of annual sustainability assurance engagements. The ICVCM's CCP framework is establishing cross-registry quality standards. None of this was part of the original design requirements for registry systems that are now two decades old.
What Fixing This Looks Like
We're not arguing that the registries need to rebuild their core systems from scratch. The issuance and retirement tracking that registries provide is functional and generally reliable. What's needed is an API-accessible data layer that connects registry issuance records to the upstream measurement and verification documentation.
The minimum viable data linkage would include: a structured reference from each issuance batch to the verification report that authorized it; a structured reference from the verification report to the monitoring data it reviewed; and a queryable interface that allows any party with a serial number to retrieve the full upstream reference chain. The documents themselves don't need to move into the registry — they can remain where they are. What needs to exist is a verifiable pointer structure that allows the chain to be traversed.
Some of this is already happening. The ICVCM's tagging program for CCP-approved credits is a step toward cross-registry quality signaling. The Article 6 machinery of the Paris Agreement is driving progress on correspondent adjustment documentation. But neither addresses the fundamental data-linkage problem at the measurement-origin level. That work remains to be done, and the market's credibility recovery depends on it moving faster than the next wave of investigative reporting.
