How to use a SaaS transparency score in procurement

A SaaS transparency score is useful when it tells you where to look next. It becomes misleading when a buyer treats it as a complete security assessment, a contract review or a recommendation. The score on this site summarises a defined set of public findings and measured behaviour. To use it in procurement, keep the number attached to its criteria, measurement coverage, date and underlying evidence.

This guide walks through that reading process. It uses the scoring rules checked on 2026-09-28, version v1, and clearly labelled hypothetical examples. The examples describe mathematical inputs, not any company in the catalogue. For current rules, use the methodology; for the meaning of an individual finding, follow its criterion page.

Start with the decision the score can support

Before comparing numbers, write down the buying question. Perhaps you need to locate public contract material before a sales call, identify questions about data handling, or prepare a shortlist for a specialist review. These are research tasks. A single transparency score cannot settle whether a service fits your architecture, your contractual requirements or the intended use of your data. Those questions need their own evidence and responsible reviewers.

Use the score to navigate the record, then read the findings that matter to your requirement. A high overall result can still leave a particular question unresolved. A lower result can still lead to a useful source for the exact document you need. The practical value lies in reducing the work needed to find and assess evidence, rather than replacing your decision with the order of a list.

Read the weights as a published choice

The criteria do not contribute equally. Their weights are part of the published method, not a discovery about the inherent importance of every requirement in every organisation. Your procurement team may have a non-negotiable condition that cannot sensibly be traded against several unrelated positive findings. Keep that condition visible even when the overall number looks attractive.

The chart below expresses each criterion weight against the full available weight. It explains the rule, rather than reporting how many companies passed. The complete list matters: looking only at the largest weights can hide a question that is important for your particular use. For example, a source relevant to incident communication and a source relevant to contract terms answer different questions, even when both contribute to one total.

When sharing a result, name the version as well as the score. A number from a changed method cannot automatically be read as movement in a vendor’s behaviour. Versioning makes that distinction visible. The developer documentation explains how to retrieve the published records and their context, so a team can preserve the evidence used in a particular review.

Measurement coverage changes the denominator

The method distinguishes a measured negative finding from a criterion that was not measured. An unmeasured criterion is excluded from both the earned weight and the available weight for the calculation. The record also needs enough measured criteria before a score is published. This prevents an operational failure in collection from being silently treated as a fact about a company.

That distinction also matters when two results are compared. Their denominators may differ. A score based on a narrower measured set answers a narrower mathematical question, even when it passes the publication threshold. Read coverage before interpreting a small difference in the displayed number. A percentage looks precise while saying little about which particular evidence is available.

The second illustrative percentage is higher because the denominator is smaller. It is not evidence of a better product. Neither example should be used as a target score for a real supplier. They simply expose the arithmetic: positive measured weight divided by all measured weight, then rounded for display. Reading the numerator alone would miss the reason for the difference.

Turn each finding into a procurement question

A public document being found is a starting point for review. You still need to check its subject, scope, date and relevance. A document might concern an organisation, a particular product or a specific contractual arrangement. Your intended purchase may involve a different plan, deployment option or contracting entity. Preserve those distinctions when moving information from a public page into a buying file.

Use a simple evidence register with a requirement, source, retrieved date, observation and open question. Write the observation narrowly: a specified page contains the relevant statement, or a particular check returned its recorded result. Then state what remains to be verified. This makes the handoff to procurement, legal or technical colleagues more useful than sending a screenshot of a score without its source trail.

Use the explorer to locate records that match the public findings you want to examine. Treat the resulting list as a research set. Conditions based on public observations should not silently become a complete specification for your purchase. If a required question remains open, assign it explicitly rather than assuming that a different positive finding compensates for it.

Handle a missing finding without inventing a verdict

A search result describes the collection process at the addresses and time checked. A missing finding does not settle every possible publication location, private document or contractual arrangement. Keep the wording attached to the measurement. In a buying file, “not found at the checked addresses” remains understandable even if a vendor later supplies a relevant document through another channel.

Ask for the evidence needed to resolve your requirement. A vague request to “improve the score” is less useful than a question about a specific document, scope or statement. When new evidence arrives, record its date and origin. A document supplied privately belongs in your authorised review process; it does not retrospectively change what was observed on a public page.

Keep the original observation and the later clarification distinguishable. A future reviewer should be able to see why the question arose and what answered it. This matters especially when a procurement decision outlives the people who first assembled the shortlist. The record should explain the conclusion without requiring someone to remember a sales conversation or find a message in another colleague’s inbox.

A handoff checklist for the buying team

1. Describe the buying requirement and any condition that needs its own approval.

2. Record the score version, date and measurement coverage.

3. Follow each relevant finding to its source and check the subject and scope.

4. Separate the observation from the conclusion your team draws.

5. Assign unresolved questions to a named owner with a specific evidence request.

6. Preserve the reviewed material and define what changes require another look.

For a handoff, include a short decision note as well as the table. State which questions have been answered, which remain open and who is responsible for the next step. Avoid a single green status for a file containing several unresolved conditions. The purpose of a concise note is to make the next review easier, not to conceal uncertainty behind a polished summary.

After a change in requirement, product version or relevant source material, revisit the affected conclusion. You need not pretend that every old observation has become useless. Identify which part still applies and which part needs new evidence. This keeps the review proportionate while preserving a clear account of the limits under which the earlier decision was made.

Questions about using the score

Method and sources

The public endpoint carried dataset identifier r133.107528 when retrieved on 2026-09-28. That identifier records the source response used for the rule check; this article does not turn the dataset into a new market survey. Keep a dated rule check separate from a later observation about an individual record.

The rule set was checked on 2026-09-28 against the public criteria endpoint and the scoring implementation. The endpoint returned score version v1. The chart of weights is calculated from that dated rule set. The examples are hypothetical inputs in criterion order, explicitly separate from vendor observations. No market distribution or named vendor result is claimed here.

curl -s https://saas-ranking.com/api/v1/criteria
curl -s https://saas-ranking.com/methodology

These commands retrieve the current public method, which may differ from the dated rules in this article. The methodology links the criteria and explains the publication threshold. For a deeper look at what particular source statements mean, read the linked criterion descriptions before using any result in a contractual or technical decision.