Score Source Journal

Published by Lana

Cookie settings

Understanding score differences

How to Compare Credit Scores Without Comparing the Wrong Things

Use five labels to compare credit scores: provider, model, version, bureau, and calculation date.

Editorial analysis. This publication cannot access accounts, retrieve scores or provide support for a credit provider. Official-source links lead away from this site.

In this article

Two different credit scores do not automatically mean that one service made a mistake. Before comparing the numbers, compare the descriptions attached to them. A useful comparison begins with the source and ends with the number, rather than the other way around.

The method here is a five-label record: provider, model family, model version, credit bureau, and calculation date. Keep any stated score range alongside those labels. This approach will not reproduce a scoring formula. It will help you recognize when you are comparing different measurements and when a discrepancy deserves a closer look.

1. Record the provider without confusing it with the model

The company displaying a score is not necessarily the company that developed the scoring model or the bureau supplying the data. A familiar bank’s app can be the display location while the score label identifies a separate model and reporting source.

Start a private note with the provider’s name and the page where the explanation appears. Look for nearby disclosures, information icons, or a help article about the score feature. Record only what the provider states. If the page does not identify the model, mark that field “not disclosed” rather than filling it in from a search result about someone else’s account.

This creates a useful question for support: “Which model, version, bureau, and calculation date apply to this score feature?” It is more precise than asking why your score is different everywhere.

2. Identify the model family and exact version

FICO and VantageScore are different scoring model families, and both have versions. The number following a model’s name matters. A brand name alone does not establish that two scores were calculated the same way.

FICO’s explanation of score versions distinguishes base scores from industry-specific versions and notes that a lender may use a different version from the one a consumer sees. VantageScore’s consumer FAQs likewise describe model versions and explain that different scoring models are not interchangeable.

Copy the full label, including an auto or bankcard designation when present. Do not shorten everything to “FICO” or “credit score.” Also record the displayed range. A color label such as good or excellent is a presentation shortcut; it should not replace the actual model information in your comparison.

3. Find the bureau behind each result

Next, identify the credit reporting source: Equifax, Experian, or TransUnion, when disclosed. A score display from one provider does not tell you that all three reports were used. Read the source statement instead of inferring it from a logo or the site’s marketing language.

The CFPB explains that reporting sources can contribute to different scores. Treat the bureau as a required comparison field, not a footnote you can ignore.

If one provider identifies a bureau and the other does not, your comparison is incomplete. You can still keep both results, but label the uncertainty. When you need to examine the underlying information, review the relevant reports rather than asking a score graphic to explain an account-level difference. Our report-versus-score guide describes that review.

4. Distinguish the calculation date from today’s date

Opening an app today does not necessarily mean the displayed score was calculated today. Look for language such as calculated, updated, or as of. Then check what that language describes. A page refresh, a monitoring alert, and a score calculation are not interchangeable events.

FICO states that its scores reflect credit file information at the time they are requested and that results can differ across agencies and dates. VantageScore’s explanation of differing scores also identifies model, bureau, and refresh timing as reasons to check the labels.

Write the date next to each score before considering a trend. If the date is unavailable, say so. Do not assume a service updates immediately after a payment, a report correction, or another event. Ask the provider about its display schedule when that timing matters to your interpretation.

5. Use a fictional comparison to test the method

Imagine that Luis sees a higher score in a bank app than in a separate consumer dashboard. His notes show that the bank displays a FICO score from one bureau, while the dashboard displays a VantageScore version from another bureau. The two calculation dates are also different.

Luis has learned something useful without deciding which number is “real.” The results differ on several important dimensions, so their gap is not a clean measurement of improvement or decline. He labels them as separate series rather than combining them into a single chart.

On his next review, he compares each source with its own earlier result, checking that its model and bureau have not changed. If a provider changes models, he starts a new comparison series. This is a recordkeeping method, not a claim that any specific trend will continue or that one model is superior.

6. Investigate mismatches without inventing a conversion

When two results appear to share the same labels, review the disclosures again. The date might refer to different events, or a version detail may be missing. Ask the provider for clarification before assuming an error. If the underlying reports show a factual discrepancy, investigate that record directly.

Avoid online rules that promise to convert one model to another by adding or subtracting a fixed amount. Your comparison note should preserve what each source actually reported, not produce a new number that no provider calculated. The same caution applies to averaging multiple scores.

If you are preparing for a credit application, you may ask the prospective lender which score information it can identify. Do not assume an educational dashboard duplicates its decision process. A score comparison cannot establish approval, terms, affordability, or the full information a lender will consider.

7. Keep the comparison useful and limited

A complete entry contains the provider, full model label, bureau, calculation date, score, and stated range. Add one sentence describing any uncertainty. That is enough for a manageable record; full report screenshots and sensitive identifying details are unnecessary for this worksheet.

Use a consistent review routine instead of repeatedly opening multiple apps in search of a preferred result. Before adding another service, inspect its terms with the free-score offer checklist. More sources can create more complexity without answering your original question.

This independent educational guide does not calculate, retrieve, validate, or translate your scores. It offers a way to organize disclosed information so that your next question goes to the right provider with the right context.

Sources and scope

Public sources checked November 4, 2026. The reading framework and fictional examples are original educational analysis.

  • FICO: Score versions: Base and industry-specific versions; date and bureau differences; lender versions may differ.
  • VantageScore: Consumer FAQs: Model versions and non-interchangeability.
  • CFPB: Understand your credit score: Different reporting sources and score models.
  • VantageScore: Why are each of my credit scores different?: Bureau, model, and refresh timing differences.

Have a public source that changes this analysis? Suggest a correction. Please don’t send private credit records, financial identifiers, identity documents or account credentials.