Credit score methodology
Know what the number means. Understand the evidence behind it. Make a credit decision you can explain.
Business Credit Reports.
Anonymised report example · United Kingdom
Active company Business support services
Credit score
Very Low Risk
Recommended limit
£1.2m£1,200,000 · GBP
Modelled assessmentHistorical snapshot. Report generated 23 June 2026.
A credit score is an assessment, not a guarantee of payment.
The 34-second overview
Follow the evidence, the assessment and the questions your credit team needs to ask next.
Start with the legal company. Review its filed financials and available payment and legal records. Keep source records separate from the modelled credit assessment. In our anonymised historical example, the score is 87 out of 100 and the recommended limit is £1.2 million, both dated 25 September 2025. The previous entry shows a score of 96 and a limit of £3.5 million. Read both histories, the financial period and the assessment date. Then consider your existing exposure, proposed terms and credit policy. The report supports the decision; your team makes it.
01 / Understand the methodology
A reliable review keeps the underlying company information, the assessment and your own policy in view.
The recorded facts
Legal identity, filed accounts, registered charges and court records establish the available evidence. Trade payment experiences add a separate view from contributing partners.
Source informationThe credit assessment
The score, risk description and recommended limit summarise the assessment. Read them with the source dates, available history and supporting commentary.
Modelled outputsYour credit policy
Set terms against your actual exposure, group concentration and risk appetite. A commercial decision also needs the context your business holds.
Your organisation’s decisionBusiness Credit Reports’ published credit-risk overview ↗ describes these headline specifications. Confirm the model, scale and available inputs for the markets in your portfolio.
02 / Inspect the evidence
Explore the information available in our anonymised example. These are review areas, not a published weighting of the scoring model.
Filed accounts & financial indicators
Review profitability alongside liquidity and capital. The latest accounts in this example cover the year ended 31 December 2024, with unconsolidated figures.
Financial statements · period-specific data
Figures transcribed from the anonymised report. Compare available years and read the accounting basis before drawing a conclusion.
Selected third-party contributors
Payment experiences describe a contributor sample. Look at the period, sample size and available ageing information before treating it as representative of a company’s obligations.
Partner-sourced information · separate from registry filings
This contributor sample does not cover the company’s full sales ledger. The view shows the reported sample size and average invoice value; it does not estimate an on-time payment rate.
Registered security & legal information
The example includes dated charge records and their supporting descriptions. In a review, distinguish a record’s existence from its current status and the exposure it may affect.
Registered records · coverage and detail vary by jurisdiction
Eight records shown does not mean eight outstanding charges. Review individual status or satisfaction information where supplied.
Legal profile & corporate relationships
Match the company on the contract to its legal identity. Review shareholders, directors and corporate relationships alongside the individual company’s finances.
Company information · distinct from a payment guarantee
A parent relationship does not by itself establish a parental guarantee or an obligation to support a subsidiary. The example’s original names and identifiers are removed.
Availability depends on the company and country. A missing field is a gap to investigate; it should not be read as evidence of low risk.
Search business credit reports
Select a country and enter the company’s legal name.
Free company preview. Full reports available on request.
Search by company number or VATEvaluate it against your own requirements
Review the markets, evidence and credit outputs your team needs before you commit.
03 / Read the movement
Both can be true. In this example, the score moved from 96 to 87 while the recommended limit fell from £3.5m to £1.2m.
Anonymised historical report
United Kingdom · Business support services
Selected assessment
Credit score
Recommended limit
£1.2m£1,200,000 · GBP
Review changes in both measures. The history alone does not establish why either changed.
25 September 2025: score 87 of 100, recommended limit £1,200,000 GBP.
| Assessment date | Credit score | Recommended limit (GBP) |
|---|---|---|
| 2021-10-13 | 90 | £1,700,000 |
| 2022-10-14 | 95 | £2,150,000 |
| 2023-10-07 | 91 | £2,350,000 |
| 2024-10-15 | 96 | £3,500,000 |
| 2025-09-25 | 87 | £1,200,000 |
The period the accounts describe.
When the accounts entered the record.
The date of the latest assessment entry.
When this report was produced.
For credit committees & procurement
Before using any score in an automated policy, review the specification and test how it behaves on your own portfolio.
Discuss a methodology evaluation →Which companies return a score? What happens when financials or payment records are missing?
What outcome and time horizon does the score assess? Which scale and model apply to each market?
Ask about performance testing, model revisions, update timing and how material changes are communicated.
Keep the dated report, approved exposure and reason for any override together.
Fit the way your team works
Company research and reports for your analysts.
Credit report data inside your own systems.
Supported company-data tools for compatible AI assistants.
Agree coverage and licensing for portfolio workflows.
Credit score methodology · FAQs
Straight answers for enterprise credit, risk and procurement teams.
A company credit score is a modelled assessment of credit risk. Business Credit Reports’ published overview describes financial information, payment behaviour, legal records and company context among the review areas. The model and available inputs can differ by market. This page explains the evidence and interpretation; it does not publish coefficients, a universal formula or model validation results.
In our anonymised UK example, the report assigns a score of 87 and the description “Very Low Risk”, with the latest history entry dated 25 September 2025. It is a historical assessment on the scale shown in that report. Read the risk description, assessment date and supporting information together. Do not transfer a numeric threshold from one country or provider to another without checking the applicable scale.
No. A score of 87 out of 100 does not mean an 87% chance of payment or a 13% chance of default. Where a report provides a separate probability measure, confirm its definition, unit and time horizon before using it. A credit score is also different from the issuer or bond rating assigned by a ratings agency.
The score summarises the assessment of credit risk. The recommended limit provides a monetary reference for exposure. In the example, the score falls from 96 to 87 while the recommended limit falls from £3.5 million to £1.2 million between the 2024 and 2025 entries. Review both alongside existing receivables, proposed orders, payment terms and your internal approval policy.
Source types should be kept distinct. Company identity, filed financial statements and registered charges can come from official records. Court information depends on local registers. Trade payment experiences come from contributing partners, rather than company accounts. Scores and recommended limits are modelled assessments. Confirm the provenance and coverage of the fields relevant to your markets.
Availability varies by company, legal form and country. A new business or an entity with limited filing obligations may have a thin financial record. Missing payment information does not establish that invoices are paid on time. Review the available evidence, identify the gaps and confirm whether a usable score is returned before incorporating it into an automated policy.
Check the scale, risk description, model and source coverage first. A number can look comparable while referring to a different scoring system or data environment. The API reference includes provider score values and scale fields, as well as risk descriptions where available. Agree the interpretation and any mapping with your team before applying portfolio-wide thresholds.
New information and model revisions can affect an assessment. Check the dated score and limit history, and read any supporting commentary. The financial period end, accounts filing, assessment entry and report generation date describe different events. A newly generated report can contain an earlier assessment. The history alone does not prove which input caused a change.
Business Credit Reports provides a Credit Report API for supported company reports, plus platform access and agreed data feeds. MCP provides supported company-data tools to compatible AI assistants and agent workflows. The available MCP tools and permissions should be checked for your use case; API report fields are not automatically available as individual MCP tools.
Review market and company coverage, missing-data treatment, score definitions, the model’s target outcome and horizon, validation evidence, update timing and change management. Test the proposed policy on your own portfolio and document overrides. Contact our team to discuss the information and evaluation needed for your requirements. A report supports your decision; it does not guarantee payment.
Make the next decision with context
See the available financials, credit assessment and supporting records for the companies that matter to your business.