Why vendor sustainability scorecards matter for martech, analytics and hosting

Martech, analytics and hosting suppliers deliver different kinds of services but share a common procurement challenge. Vendors often present sustainability claims that are hard to compare because they report different metrics, cover different boundaries and mix product improvements with corporate pledges. A structured scorecard turns varied disclosures into comparable answers so teams can choose providers that align with product goals, compliance requirements and operational carbon priorities.

What a good scorecard accomplishes

Consistent comparison of emissions and energy related controls across providers. Actionable requirements procurement can use in contracts and SLRs. Clear signals to vendors about which product features and disclosures matter to customers.

Scope and boundaries to include for martech, analytics and hosting

Start by defining which parts of vendor activity the scorecard covers. For hosting the focus is often on data center energy and purchased electricity. For analytics and martech SaaS the relevant items include multi tenant compute, data storage, event collection volumes and any customer facing compute such as model inference. All vendor assessments should reference greenhouse gas reporting boundaries using established categories rather than ad hoc labels.

Standard reporting frames to require

Request information aligned to established frameworks rather than bespoke summaries. Key references are corporate reporting for Scope 1, Scope 2 and Scope 3 as defined by the Greenhouse Gas Protocol. Where vendors report targets, check whether those are set through recognised programs such as the Science Based Targets initiative. For operational controls include certifications such as ISO 14001 for environmental management when they exist.

Metrics and evidence to request

Collect both quantitative metrics and qualitative evidence. Quantitative metrics enable scoring and normalization. Qualitative evidence explains assumptions and control maturity.

Quantitative metrics to ask for

1. Total annual Scope 1 and Scope 2 emissions in metric tons CO2e and the reporting year.
2. Scope 3 categories that are material to the service such as purchased IT hardware, upstream cloud provider emissions when applicable and business travel; specify which categories are included.
3. Energy use or fuel consumption data where available for hosting providers and owned data centers.
4. Emissions or energy intensity metrics normalized to useful service units such as emissions per TB stored, per million events processed, per 1 000 active accounts or per model inference hour depending on the product type.
5. Renewable procurement details: power purchase agreements, renewable energy certificates and whether these are regionally matched to consumption.
6. Data retention and average event payload sizes for analytics and martech tools where they affect data transfer volumes.

Qualitative evidence to request

1. Copies of third party assurance reports or verification statements linked to emissions disclosures.
2. Dates and scopes of any environmental management certifications.
3. Technical controls that reduce compute or transfer such as configurable sampling, server side batching, configurable retention and edge delivery options.
4. Design commitments such as energy efficient inference, model size limits or client side processing options that reduce server load.
5. Policies or contract clauses that enable regional data placement and scheduling to avoid high grid carbon intensity periods.

Designing the scorecard: structure and weighting

A practical scorecard balances governance and operational controls with measured outcomes. Governance and disclosure reduce risk and improve transparency. Operational metrics and product features determine actual service impact.

Suggested assessment categories

Use a concise set of categories so procurement can make decisions quickly. Typical categories are governance and disclosure, emissions intensity, renewable procurement, product and engineering controls, and verification and contractual readiness. Allocate weights to reflect your priorities. For example, a regulated organisation may weight verification and contractual readiness more heavily while a product team optimising for emissions per request may weight product controls higher.

Scoring approach

Translate each vendor answer into a score using clear rubrics. For disclosure require reported Scope 1 and 2 data as a minimum. For intensity metrics accept vendor reported normalised values but require documentation of numerator and denominator. For product features score presence of controls, configurability and measurable outcomes. Keep thresholds and score boundaries documented so vendors understand how to improve over time.

Normalization: how to compare different service types

Directly comparing a hosting provider with a martech vendor requires careful unit choices. Where possible normalise emissions to service units that reflect how your teams use the product. For example, emissions per gigabyte stored suits storage services. For analytics platforms emissions per million events or per hour of query time can be useful. If vendors cannot provide per service unit metrics use proxy normalisation such as emissions per revenue adjusted for typical cost per unit, while noting the limitation.

Handling unknowns

If a vendor cannot provide a requested metric, score that absence but allow secondary evidence such as engineering descriptions or partial measurements. Encourage vendors to improve data availability by making clear which metrics will be required in future procurement cycles.

Verification and red flags

Not all disclosures carry the same reliability. Third party assurance and standardised reporting indicate higher confidence. Red flags include consolidated corporate level emissions that cannot be traced to the product you will consume, unverifiable renewable claims, inconsistent reporting periods and metrics that change without explanation.

Third party checks to prefer

Independent assurance statements, participation in recognised disclosure platforms such as CDP and certification against environmental management standards provide stronger evidence than unaudited slides. For hosting services ask for data center level PUE measurements and clarifications about whether renewable procurement is regionally matched to consumption or only a global offset.

Operationalising the scorecard inside procurement and product teams

Integrate the scorecard into each stage of vendor selection and contract management. Use the scorecard to shortlist vendors, to inform negotiation levers and to define contractual reporting obligations. Make the scorecard part of regular vendor reviews and require remediation plans where vendors score poorly on material categories.

Implementation steps

1. Define the mandatory evidence and the aspirational evidence you will accept.
2. Run a pilot with a small set of vendors to refine questions and scoring thresholds.
3. Publish the scorecard criteria to vendors so they can prepare evidence in advance.
4. Include scoring outcomes in procurement decisions and track vendor improvements over time.

Sample vendor questionnaire items for martech and analytics

Below are concise items you can include in requests for information. Tailor units to how your teams consume the service.

1. Provide Scope 1 and Scope 2 emissions for the last reporting year and the boundary used to calculate them.
2. Provide any Scope 3 categories material to this service and the methodology used to estimate them.
3. Provide emissions intensity figures normalised per agreed service unit and document numerator and denominator definitions.
4. List product features that reduce data transfer or compute such as configurable sampling, server side batching, client side processing or retention controls.
5. Describe renewable energy procurement and whether purchases are regionally matched to consumption.
6. Attach any third party assurance or certification documentation relevant to the disclosures above.

Using scorecard results responsibly

Scorecards are decision support tools not absolute truths. Use them to prioritise vendors, to create minimum contractual requirements and to set improvement milestones. When a vendor lacks mature disclosures you can require time bound remediation and specify reporting cadence. Where vendors make public claims require substantiation or avoid using those claims to justify procurement unless verified.

Disclosure language to request

Ask vendors to provide the methods behind any numbers and to state uncertainty or estimation methods. Request that any offset or renewable claims are described in terms of instrument type and regional matching. Clear language reduces the risk of accidental greenwashing and makes future audits simpler.

How to evolve the scorecard over time

Start with a minimum viable scorecard focused on disclosure, key intensity metrics and product controls that your teams can act on. After two procurement cycles revisit category weights and required evidence based on vendor responses and internal priorities. As vendor reporting improves you can raise the floor for acceptable answers and move more emphasis onto product level intensity and operational performance rather than corporate pledges.

Embedding a transparent, repeatable scorecard aligns procurement choices with engineering goals and reduces the risk of selecting suppliers whose reported sustainability does not map to the product impact you will actually create.