What a vendor sustainability scorecard does and who should use it

A vendor sustainability scorecard turns qualitative sustainability claims into comparable, auditable signals you can use in procurement, contract renewals and platform governance. Product owners, procurement teams, site reliability engineers and marketing operations professionals can all use the same scorecard template to compare martech, analytics and hosting suppliers against a shared set of expectations.

Core measurement categories every scorecard should cover

The scorecard should capture four high level categories. Treat each category as a distinct question you expect the vendor to answer with measurable evidence.

• Emissions and intensity. Ask for absolute emissions and activity based intensity metrics that relate to the service you buy.
• Energy sourcing and grid context. Capture whether energy is from contracted renewables, residual mixes, and the region level carbon intensity that affects delivered emissions.
• Operational efficiency. Measure how efficiently the service uses compute, storage and network resources for a unit of work.
• Transparency and verification. Require documentation, third party assurance and a published methodology for the footprint figures.

Why each category matters

Emissions and intensity let you compare the climate impact of different suppliers for the same workload. Energy sourcing determines how much the vendor can reliably reduce carbon through procurement. Operational efficiency drives the actual per request or per gigabyte emissions you will cause. Transparency is the control that lets you trust the numbers you use to make decisions.

Metrics and evidence to request by vendor type

Different vendor categories expose different practical metrics. Below are the concrete items to ask for and the form the evidence should take.

Martech vendors and tag servers

Request metrics that translate to per page view or per request impact. Practical asks include a measured average data transfer per request, average execution time of client side scripts in milliseconds, and any server side compute used per 1000 events. Ask for a clear method showing how those figures were measured and the test workloads used. Evidence can be a lab measurement report, a client side performance trace, or a reproducible script you can run against a demo account.

Analytics vendors

Ask for the vendor footprint of data ingestion, storage and query processing per million events or per gigabyte stored. Request data transfer volumes for a typical API call and the distribution of storage classes used. Ask whether the provider applies aggregation or sampling that reduces processing. Evidence should include a published methodology, recent emissions inventory and, where available, third party assurance of the inventory.

Hosting and infrastructure

Require the hosting provider to supply data centre level metrics such as power usage effectiveness and regional carbon intensity for the electricity supplied to services you will use. Ask for the provider specific emissions per kWh or per VM instance type and any applied renewable energy purchases or power purchase agreements. Evidence should include certificates for energy procurement, ISO 14001 or ISO 50001 registrations if held, and any independent audit reports that attest to the provider emissions claim.

Practical units and normalisations to request

To compare vendors fairly you must normalise their figures to a consistent unit. For martech and analytics the most useful normalisations are per 1 000 page views, per 1 000 events, and per gigabyte transferred. For hosting the natural normalisations are per kWh, per VM hour, and per terabyte transferred. Require vendors to publish how they convert raw electricity use to greenhouse gas emissions and to state the emission factor source and date used.

Sample scoring framework you can adapt

The scorecard should map measurable evidence to a numeric result you can aggregate. The following framework is a practical starting point you can adapt to your risk appetite and procurement policy.

1. Transparency weight. Score whether the vendor provides a published methodology, raw data on request, and third party assurance. Higher scores for independent verification and machine readable reports.
2. Operational intensity weight. Score the measured per unit impact. Lower measured impact means a better score. Use your normalisation units to compare vendors directly.
3. Energy sourcing weight. Score the proportion of electricity backed by contracted renewable supply and the regional carbon intensity. Prefer vendors who can demonstrate contracted or additional renewable energy for the services you use.
4. Improvement trajectory weight. Score whether the vendor has a time bound plan to lower emissions, with milestones and a published roadmap.

Assign weights that reflect organisational priorities. For example where procurement policy prioritises verified transparency you can give transparency a higher weight. Use a 0 to 100 final score to keep comparisons simple and store both the aggregate score and the underlying evidence links in your vendor registry.

Verification and evidence standards to require

Insist on evidence that aligns to widely recognised standards. The Greenhouse Gas Protocol is the common framework for defining scopes and inventory methods. For energy management look for ISO 50001 registrations and for environmental management ISO 14001 registrations. Where vendors claim low carbon electricity use ask for contractual evidence such as renewable energy certificates, power purchase agreement summaries or supplier invoices that match the reporting period. If a vendor publishes a footprint, require either a statement of assurance from a qualified assurance provider or a clear explanation of why assurance is not performed yet and when it will be available.

How to integrate the scorecard into procurement workflows

Embed the scorecard into your RFP template and require vendors to complete it as a mandatory annex. During evaluations run a light verification step that checks the supplied documents against published reports and sample measurements. If a vendor scores below your threshold allow a remediation period where they can submit additional evidence or propose compensating measures such as a plan to migrate workloads to lower impact storage tiers. Reassess the vendor at contract renewal and attach performance related clauses that require updates to the scorecard on a fixed cadence.

Contract language and operational clauses to include

Include clauses that require the vendor to update their reported figures annually, to notify you of material changes to energy sourcing or infrastructure, and to cooperate in on demand measurement tests. If your organisation uses carbon budgets or internal shadow pricing include language that allows you to estimate expected costs from projected emissions. Require the right to request and reproduce measurement scripts or sample telemetry that proves the performance claims made during procurement.

Common pitfalls and how to avoid them

One common pitfall is accepting headline claims without underlying data. Vendors often publish statements such as reduced emissions or 100 percent renewable energy. Ask for the precise scope covered by those claims. Another pitfall is comparing absolute emissions across vendors without normalising for the workload type. Normalise to per request or per gigabyte to get an apples to apples comparison. Finally be wary of double counting when vendors use unbundled renewable certificates to claim reductions. Prefer contractual instruments that demonstrate additionality or long term renewable procurement for the grid region you operate in.

Example scorecard fields to collect

Collect the following fields as a minimum. This set ensures you can reproduce the vendor score and rerun calculations if factors change.

• Reported scope 1 and scope 2 emissions for the last 12 months and the methodological boundary used.
• Reported scope 3 categories relevant to the service and how they were estimated.
• Per normalised unit intensity such as gCO2e per 1000 page views or per 1000 events.
• Energy procurement evidence including certificates or contracts and the period covered.
• Data centre PUE or equivalent efficiency metrics where applicable.
• Third party assurance statements or links to audited reports.
• Published roadmap with milestones and target dates for emission reductions.

Questions to ask in an RFP to surface reliable answers

Use a short list of direct questions that force vendors to provide evidence rather than marketing language. Ask them to provide the exact emission factors and sources they used for each calculation. Ask for the measurement scripts or command lines that reproduce any lab measurements. Ask for the auditor name and scope of any assurance and for copies of certificates for renewable energy procurements. Finally ask them to describe the normalisation used to convert raw usage into the per unit figures they supply.

Next steps for teams building a scorecard

Start by piloting the scorecard with two or three vendors in different categories. Use the pilot to refine weights and to tune the normalisation units for your traffic and workloads. Store the scorecard results alongside contractual documents and set a recurring review to ensure scores reflect current operations. Over time you can add additional layers such as social and circular economy criteria, but keep the initial scorecard focused on measurable greenhouse gas impacts and verifiable evidence.