Hard Inquiry

A hard inquiry is an entry on your credit report showing that a lender or other permitted user obtained the report in connection with an application you made for credit. Fair Isaac's description is the working one: "Hard inquiries such as actively applying for a new credit card or mortgage may affect your score," and they are "visible to lenders who view your credit reports."

It is worth being precise about where these words come from, because a great deal of writing on this subject implies a statute that does not exist. The phrases "hard inquiry" and "soft inquiry" appear nowhere in the Fair Credit Reporting Act. They are trade and scoring-company vocabulary. The statute does draw a related distinction, but it draws it on visibility rather than on application: under 15 USC 1681b(c)(3) a consumer reporting agency may not furnish to any person a record of inquiries made in connection with a credit or insurance transaction that the consumer did not initiate. That is the legal reason a promotional inquiry is invisible to lenders, and the trade distinction sits on top of it rather than being defined by it. The two lines are close but not identical, which is why an account review by an existing creditor is not consumer-initiated and is nonetheless treated as soft.

How much it costs, in Fair Isaac's own words. "For most people, one additional credit inquiry will take less than five points off their FICO Scores." Two qualifications come with that from the same source. "Inquiries can have a greater impact if you have few accounts or a short credit history," so the effect is largest exactly where a borrower can least afford it. And the effect fades: "Hard inquiries stay on the report for up to two years, but they only affect the FICO Scores for a year."

The 10 percent belongs to a category, not to inquiries. Fair Isaac says inquiries "play a minor part in only 10% of what makes up a FICO Score," and that 10 percent is the new credit category, which has three inputs: how many new accounts you have by type of account, how many recent inquiries you have, and how long it has been since you opened your newest account. So writing that inquiries are 10 percent of a score overstates them by an unknown factor. The weights across all five categories belong to the FICO Score page.

Rate shopping is the most misreported part of this subject, and the fix is to keep two separate mechanisms separate.

The first is grouping. Fair Isaac states that "FICO Scores group multiple hard inquiries that are made within a short time frame (14 to 45 days) into one inquiry," and explains the range rather than leaving it as a hedge: "For FICO Scores calculated from older versions of the scoring formula, this shopping period is any 14-day span. For FICO Scores calculated from the newest versions of the scoring formula, this shopping period is any 45-day span. Each lender chooses which version of the FICO scoring formula it wants the credit reporting agency to use." The 14 and the 45 are therefore properties of the model version the lender happens to order, not alternatives a borrower selects, and averaging them into "about a month" throws away the only information that makes the range usable.

The second is a buffer at the moment of scoring, and it is a different operation. "For loans that commonly involve rate-shopping, such as mortgage, auto and student loans, FICO Scores ignore inquiries made in the 30 days prior to scoring." Grouping collapses several inquiries into one; the buffer disregards recent ones entirely while the shopping is still in progress.

Two boundaries on that, both of which consumer articles routinely cross. Fair Isaac states the buffer for the loan types it names, so extending it to credit card applications is not supported by what the company publishes. And the grouping is described for "numerous hard inquiries of the same type (student, auto, or mortgage loans)." What follows is a careful negative rather than a confident one: Fair Isaac publishes no equivalent treatment for card applications, so several card applications in one week cannot be assumed to be grouped. Asserting that they are never grouped would be a claim about a proprietary model that nobody outside the vendor can check.

Where the law does bear on inquiries. Two provisions are worth knowing. Under 15 USC 1681g(a)(3)(A), on request a consumer reporting agency must identify each person that procured a consumer report on you, for employment purposes during the preceding two years and for any other purpose during the preceding one year. So the record of who pulled your file is yours to see, with a longer window for employment. And under 15 USC 1681c(d)(2), where a report contains a credit score or other risk predictor, the agency must include "a clear and conspicuous statement that a key factor" adversely affecting that score "was the number of enquiries, if such a predictor was in fact a key factor." Congress legislated on the assumption that inquiries move scores, and the statute's own spelling of the word is a small reminder of how old the provision is.

On removing one. A furnisher's and an agency's duties under the Fair Credit Reporting Act run to accuracy, so an inquiry that accurately records an application you made is not a dispute matter. An inquiry from a company you never applied to is a different case, and it can be disputed with the agency that reported it. Fair Isaac's own advice on unexpected inquiries is to check all three reports, research the company listed, and dispute what you do not recognize, since an unrecognized hard inquiry can be the first visible sign of an application made in your name.

You apply for credit and authorize the lender to obtain your report. The lender requests it from one or more bureaus, and each bureau records that request in your file. A scoring model reads the file, including the recent inquiries, and returns a number. Nothing about the inquiry changes after it is recorded; what changes is how the model treats it as it ages.

A hypothetical example of the two clocks. Suppose a lender pulls Nadia's Experian report on March 1, 2026, because she applied for a card. FICO Scores consider inquiries from the last 12 months, so that inquiry stops affecting her FICO Scores from around March 2027. It remains visible on the report for up to two years, so it is still there to be seen until around March 2028, doing nothing.

A second hypothetical, on shopping windows. Nadia applies to four mortgage lenders, the first on June 3 and the last on June 15, a 12-day span. That fits inside a 14-day window and comfortably inside a 45-day one, so on either version of the formula the four inquiries are grouped as one. Had the last application gone in on June 19 instead, a 16-day span, the set would fall outside a 14-day window while still fitting a 45-day one, and which treatment she gets depends on the version the lender ordered rather than on anything she can control. Separately, because mortgage loans are one of the types Fair Isaac names, inquiries in the 30 days before scoring are ignored, so finding a loan quickly keeps the shopping out of the score used to price it.

What the arithmetic deliberately does not do is add up point costs. Fair Isaac says one additional inquiry costs most people less than five points and that the effect varies with the file, and multiplying that figure by a number of applications would be inventing precision the company does not publish.

Pros

• The record is a genuine risk signal, which is part of why a lender who has never met you can price a loan in minutes.
• The effect is small for most files, temporary, and gone from the score after twelve months.
• Rate shopping for a mortgage, auto or student loan is protected by both grouping and a 30-day buffer, so comparing lenders is not penalized.
• You can see every inquiry, and federal law entitles you to be told who procured a report on you, going back two years for employment purposes.

Cons

• The effect is largest for people with few accounts or a short history, which is the group most likely to be applying.
• Whether your shopping fits a 14-day or a 45-day window depends on the model version the lender ordered, which you generally will not know.
• Fair Isaac publishes the shopping protections for mortgage, auto and student loans and not for credit cards, so multiple card applications carry no documented grouping.
• An accurate inquiry cannot be disputed away, and it stays visible for a year after it stops mattering.
• An inquiry you do not recognize may be the earliest sign of an application made in your name.