Credit scores like FICO and VantageScore are calculated with proprietary formulas that the scoring companies don't fully disclose. But both companies, along with the major credit bureaus, have publicly described the general categories of information that matter and roughly how much each tends to count. Understanding those categories — even without the exact math — is enough to know where to focus if you're trying to build or repair credit.
It helps to start with what a credit score is actually trying to predict: the likelihood that you'll repay a debt as agreed over the next couple of years. Every factor below exists because it's historically correlated with that outcome across millions of consumer files, not because any one factor is a moral judgment about you as a borrower.
Whether you've paid your bills on time is usually the single biggest input, commonly cited at around 35% of the general weighting used in consumer credit education. Late payments, collections, and bankruptcies can have an outsized negative effect, especially recent ones. A single 30-day-late payment on an otherwise clean file tends to matter more, proportionally, than being 30 days late on a file that already has several delinquencies — scoring models generally weigh new negative information more heavily against a track record of consistency.
What counts as "payment history" is broader than most people expect. It includes credit cards and loans, but late or missed payments that get reported to a bureau by a landlord, utility, or medical provider can also show up here if that creditor reports to the bureaus (many smaller creditors don't, which is why a missed rent payment doesn't always show up the way a missed credit card payment does).
This compares your credit card balances to your credit limits. Lower utilization is generally viewed favorably; many educators cite keeping utilization under 30% (and ideally lower, sometimes cited in the single digits for the strongest outcomes) as a common guideline, though the exact thresholds in real scoring models aren't public. Utilization is typically evaluated both per card and across all your revolving accounts combined, so a single maxed-out card can matter even when your overall percentage looks fine.
This is generally considered the second-largest factor category, around 30% in commonly cited general weightings, and it's the one most people can influence fastest — see our Credit Utilization, Explained guide for the full mechanics.
How long your accounts have been open, including your oldest account and the average age of all accounts, commonly cited at around 15% of the general weighting. This factor is slow to change — it simply takes time, and it's one reason financial educators often advise against closing your oldest card even if you no longer use it regularly, since doing so can eventually shorten your average account age once that closed account ages off your report.
Opening several new accounts in a short period, or having multiple hard inquiries, can signal higher risk to some models, though a single inquiry typically has a small effect, commonly cited at around 10% of the general weighting for this category as a whole. A flurry of applications in a short window (sometimes called "credit-seeking behavior") tends to draw more scrutiny than one isolated application for a mortgage or auto loan, and many scoring models have specific rate-shopping windows that treat multiple inquiries for the same type of loan within a short period as a single inquiry.
Having a mix of account types — revolving accounts like credit cards alongside installment accounts like auto loans, student loans, or mortgages — can be viewed positively, though this is usually one of the smaller factors, commonly cited at around 10% of the general weighting. This is not a reason to take out a loan you don't need; the effect is modest, and it's generally considered the least important lever compared to payment history and utilization.
FICO and VantageScore treat their scoring algorithms as trade secrets, and for good reason from a business standpoint: if the precise formula were public, it would be easier to game in ways that don't actually reflect lower credit risk. What is public is the general shape of the model — the factor categories above and their rough relative importance — which is exactly what this simulator and this guide are built around. Nothing here should be read as the literal formula.
Most people talk about "my credit score" as if there's one number. In practice, there are several score brands (FICO has multiple versions still in active use, and VantageScore has its own separate model line), and each bureau can produce a different result for the same person because Equifax, Experian, and TransUnion don't always hold identical data on you. A late payment reported to one bureau but not another, or an account that only shows up on two of the three files, is enough to create a real gap between the numbers you'd see from different free-score apps.
This is also why a number you see on a free banking app can differ from what a mortgage lender pulls: lenders in certain industries often use industry-specific score versions tuned for that type of credit decision, which weigh the same underlying factor categories slightly differently than the general-purpose score shown on a typical consumer app.
Income, employment status, savings balances, and where you live are not part of the FICO or VantageScore factor categories above, even though they're commonly and understandably assumed to matter. Lenders may look at those things separately when deciding whether to approve you and at what rate, but they are not inputs to the credit score itself. Checking your own score also does not lower it — that kind of check is a "soft inquiry" and is treated differently from the "hard inquiry" that happens when you formally apply for new credit.