Every model reads the same underlying file and asks five kinds of question. Payment history is whether past obligations were paid as agreed, and late payments, collections, and other derogatory marks weigh heaviest of all; a single 30-day late can dent a good score meaningfully, and recent problems count for more than old ones. Amounts owed is dominated by credit utilization, your card balances measured against their limits, which is the fastest-moving input in either direction. Length of credit history covers the age of your oldest and newest accounts and the average of them, which is why closing an old card can sting. New credit counts recent applications and freshly opened accounts. Credit mix looks at whether you have handled both revolving and installment debt. Fair Isaac publishes approximate weights for these five categories in its own models, along with a caveat worth carrying: the percentages describe how the categories rank for a typical profile, and their real importance differs from one person to another.
The common consumer scales run 300 to 850, where a higher number means lower predicted risk, and above roughly 740 most lenders already offer their best pricing, so chasing the last 60 points is vanity rather than value. What surprises people is that you do not have a score. Several models are in use at once, each bureau holds a slightly different file, and the number showing in a free app is frequently not the one an underwriter will price your loan against, which is worth understanding before you shop for a mortgage rather than during.
Improving a score is unglamorous: pay every bill on time (autopay the minimum as a failsafe), keep utilization low, both overall and per card, keep old accounts open, and space out applications. Checking your own score or report is a soft inquiry and never hurts. The reports themselves are free at AnnualCreditReport.com, the only federally authorized source, with one free report from each bureau guaranteed every twelve months and more frequent access currently offered as a matter of bureau policy rather than statutory right, and reviewing them for errors and unfamiliar accounts is the part of credit hygiene with the best effort-to-payoff ratio, since report errors are common and disputable.