Why FICO Scoring Factors and Myths Matter

Sam There's a number that follows you into every loan application, every apartment lease, sometimes even a job offer — and most of us only vaguely understand where it comes from. So let's start there. Maya, when a lender pulls up my credit score, what are they actually looking at?

Maya They're looking at a figure generated by a scoring company, not by the bureau itself. Nearly every score traces back to files kept by three national credit bureaus — Equifax, Experian and TransUnion — but the score is calculated by a separate company running a formula against that bureau data. The dominant one is Fair Isaac Corporation, known as FICO, used in over 90% of U.S. lending decisions.

Sam That's a layer I didn't know existed. What's inside that file they're scoring, though — is it just bills and balances, or does it pull in things like my job or where I live?

Maya Almost nothing about who you are as a person. Income, age, where you live — none of that factors in. The score comes from your credit report, which tracks borrowing and repayment behavior, and FICO scores themselves run from 300 to 850, built from five weighted categories.

Sam Five categories carrying different weights sounds like the kind of thing people guess at rather than know. I'd have assumed whether you pay on time matters most, but I wouldn't have put a number on it. What's actually driving the biggest slice of that score?

Maya Payment history is exactly that biggest slice, at 35%, and FICO's own data describes it as the strongest predictor of whether someone will pay debts as agreed. The cost of getting it wrong is steep too — a single 30-day late payment can drop a high score by 100 points or more, and it stays on your report for seven years.

Sam Seven years for one missed payment is a long memory for a system that claims not to care who you are. That feels disproportionate to a single bad month. What's the second-biggest piece, and does it carry anywhere near that kind of lasting damage?

Maya That's amounts owed, worth 30%, and it's really about credit utilization — how much of your available revolving credit, meaning cards and lines rather than fixed loans, you're currently using. Guidance converges on keeping that ratio under 30%, with scores improving further below 10%. The Department of Defense's financial program puts the truly ideal range at just 1 to 10%.

Sam Thirty percent sounds almost achievable, but one to ten is a narrow target to hit every month without obsessing over it. I'd guess most people are running well above that without realizing.

Maya They are — the average U.S. utilization ratio was 29% in the third quarter of 2024, per Experian. What makes this factor different from payment history is speed: it updates with each statement cycle, so a change in your balance shows up in your score within about 30 days, instead of lingering for years like a late payment does.

Sam So this one's forgiving in a way the late-payment damage isn't — pay down the balance and the number can recover almost immediately, rather than waiting out a seven-year sentence. That's a meaningfully different kind of factor to manage day to day.

Sam Moving on from utilization, what about how long you've had your accounts? Does the age of a card or a loan count for anything on its own, even one you hardly use?

Maya That's length of credit history, worth 15%, covering the age of your oldest account, your newest account, and the average age across all of them. Longer history generally means a higher score, so closing an old account can work against you by dragging down that average.

Sam A card sitting in a drawer with nothing on it still earning its keep feels counterintuitive, but I'll take the math over instinct. What about opening something new — does applying for a card itself cost you anything?

Maya New credit is worth 10%, and it tracks both newly opened accounts and hard inquiries — the kind that happen when you apply for credit and a lender checks your file. Multiple inquiries bunched in a short window read as risk, though the damage doesn't last: a hard inquiry can ding your score for up to six months.

Sam Six months is a short sentence compared to seven years for a late payment. Does applying for several things at once — a card, a car loan, an apartment — stack all those dings together, or does each one fade on its own clock?

Maya Each inquiry carries its own timer, but bunched together they can still pull the score down noticeably at once. The last factor is credit mix, also 10%, which looks at whether you're handling revolving credit like cards alongside installment loans like mortgages or auto loans — lenders want to see you manage different kinds of debt at the same time.

Sam Mortgages and credit cards being graded as a matched set is new to me — I'd have assumed a mortgage was just in a different universe from a credit card balance.

Maya It isn't, though it's a smaller slice of the formula than the two giants. Payment history and utilization together make up 65% of the score, which is why they're the two factors worth putting the most energy into.

Sam Here's something that's been nagging at me — if FICO and some other scoring company are both working off the same five categories, why would anyone ever end up with two different numbers?

Maya Because the two companies weight those shared categories differently. Under FICO 8, payment history is 35%; under VantageScore 4.0, it's closer to 40%, described as "extremely influential." Utilization runs the other way — 30% under FICO, roughly 20% under VantageScore. Length of history and credit mix get folded together into one combined measure worth about 21% under VantageScore, inquiries drop to roughly 5%, and VantageScore adds a factor FICO doesn't score at all, called available credit, worth about 3%.

Sam So two formulas looking at the same five ingredients can bake genuinely different cakes out of them. Is that the whole explanation for why my score might read differently depending on which bureau pulled it, or is something else going on underneath?

Maya Weighting is only half of it. The other cause is that each bureau may simply hold different data on you, since not every lender reports to all three, and reporting dates don't always line up — so a score pulled from one bureau's file can differ from another's even when it's run through the identical formula, just because the underlying report isn't identical.

Sam There's something I didn't realize until just now — that it isn't one FICO score at all, but a whole family of them depending on who's asking. What's behind that split?

Maya Mortgage lenders tend to rely on older versions, things like FICO 2, 4, or 5, while a credit card issuer is more likely pulling FICO 8. The same person can hold a dozen legitimately different scores at once, each one answering a slightly different question for a different kind of lender, and none of them is the wrong answer.

Sam A dozen versions of me floating around out there is a strange thought. Does the 300 to 850 range at least stay fixed across all of them, or does even that shift depending on the version?

Maya The scale itself holds steady at 300 to 850, and at the very top, 800 to 850 is called the Exceptional tier — that's where borrowers get the lowest rates and the best terms lenders offer. VantageScore uses that identical 300 to 850 range, but its tier boundaries don't line up with FICO's, which adds one more layer to the "why don't my scores match" puzzle.

Sam Tier boundaries not lining up sounds like a small thing on paper, but if someone's sitting right near a cutoff, that could be the difference between one offer and a noticeably worse one.

Maya It can be, which is part of why lenders rarely treat a single score as gospel. Turning to how long damage sits on your file — a late payment's sting is worst right after it happens and fades with time, but the seven-year mark on your report doesn't move regardless of how much the score impact has already faded.

Sam That's an odd mismatch — the number recovers faster than the record disappears. Does utilization work the same way, lingering on paper long after the score has already bounced back?

Maya No, utilization has essentially no memory at all. It resets every billing cycle, so a change shows up within about 30 days — and there's a nuance the Consumer Financial Protection Bureau has flagged: paying down a balance before your statement closing date, not just before the due date, is what triggers that improvement, since the statement date is when the balance gets reported.

Sam People probably assume the due date is the one that matters, since that's the one with a penalty attached if you miss it.

Maya Right, and that's precisely the mismatch worth correcting — the due date avoids a late fee, but the statement closing date is what gets reported to the bureaus. Pay down before that closing date and the lower balance is what shows up, rather than waiting for next month's cycle to catch up.

Sam And inquiries — the hard pulls from applying for new credit — those fade within something like six months, if I'm remembering that correctly from the new-credit factor.

Maya That's the window, and it's worth holding onto against a common fear: that rate-shopping for a mortgage or a car loan will tank a score for years. It won't.

Maya Closing your oldest card is a move that feels responsible, but the data runs the opposite way — shorter average account age pulls down that fifteen-percent history factor, and losing that card's credit limit also shrinks your total available credit, which raises your utilization ratio. One click, two separate hits to the score.

Sam I'd have guessed the opposite myth is just as common — that carrying a balance month to month somehow builds your score, like it's proof you can handle debt responsibly.

Maya It proves nothing of the kind. Carrying a balance just means paying interest with no scoring upside attached to it — the interest charge and the score are entirely separate transactions that happen to share a bill.

Sam So the entire "carrying debt shows lenders you're reliable" idea is just folklore with no mechanism behind it at all. On the regulatory side, is anything changing in how mortgages get scored?

Maya FICO announced newer models, FICO 10 and FICO 10T, back in 2020, and FHFA — the regulator overseeing the mortgage giants Fannie Mae and Freddie Mac — moved to require their use for conforming mortgages, with a start date in late 2025. The "T" stands for trended data, meaning the model looks at your balance and payment behavior over roughly two years rather than a single snapshot.

Sam A two-year trend line sounds like it would reward someone who's climbing out of debt, even if their balance today still looks a little rough around the edges.

Maya That's the intent — someone steadily paying down debt looks better under trended data than under a snapshot model, even with middling current utilization. And this mandate does more than swap formulas: it brings VantageScore 4.0 into the conforming-loan pipeline alongside FICO 10T, ending FICO's long-standing solo hold on that part of the mortgage market.

Sam That's a bigger shift than I expected from what sounded like a technical upgrade — a second scoring company getting written into the rules for government-backed mortgages for the first time. Before we wrap up, if someone remembers nothing else from everything we've covered today, what's the one idea you'd want stuck in their head?

Maya That a credit score is a narrow, statistically built forecast of repayment risk, not a verdict on your character or your income — it's built entirely from your credit report, and payment history and utilization alone drive nearly two-thirds of it. Different versions, different bureaus, different numbers isn't a glitch; it's just what a probabilistic system calculating the same risk from slightly different angles looks like.

Sam That's a genuinely reassuring way to think about a number that follows you everywhere. Thanks for walking through all of this with me today — it's made a pretty opaque system feel a lot less mysterious.

Maya I should say plainly before we go that this show is AI-generated, the voices you're hearing are synthetic, and every fact came from the source material named in the closing note, whose authors have no connection to this production.

Sam Appreciate you saying that, and appreciate everyone listening in — take care of that three-digit number, and we'll catch you next time.

Maya Everything today came from reporting and guidance published by Experian, myFICO, NerdWallet, Equifax, the Consumer Financial Protection Bureau, and Wikipedia, and we're grateful for the work behind all of it.