How Credit Scores Actually Work

What the score doesn't know

Your credit score doesn't know your salary. It doesn't know your savings, your degree, your job title, or your net worth. A millionaire can score 620 and a schoolteacher 810 — and the system is working as designed when they do, because the score measures exactly one thing: how you handle borrowed money. Not how much money you have. Not how virtuous you are. One narrow question, answered from one narrow file (Your Credit Report and Score's biography), by a published formula.

That narrowness is the most useful fact in this entire track, because a narrow formula is a steerable formula. This chapter lays out the recipe and the two readings of it that matter more than the percentages.

The recipe

FICO publishes the approximate weights, and they've been stable for years:¹

Payment history — about 35%. Did you pay on time, every time? The heaviest factor by design: past repayment predicts future repayment better than anything else in the file. It's also the most asymmetric — years of on-time payments build slowly, while a single 30-day-late mark can outweigh much of that record at a stroke.

Amounts owed — about 30%. Dominated by utilization: your balances against your limits, per card and overall (Credit Utilization Mechanics and Utilization Deep-Dive give the full mechanics). Not how much you owe in dollars — how much you owe relative to what you could.

Length of history — about 15%. The age of your accounts, average and oldest. This is why your oldest card is quietly one of your most valuable financial possessions (When to Close a Card (Rarely) and When Not To), doing nothing but existing and aging, like a good cellar.

New credit — about 10%. Recent applications and freshly opened accounts. A flurry of new credit reads as appetite — a mild, temporary flag (Hard vs. Soft Pulls; Rate-Shopping Windows covers the mechanics and the myths).

Credit mix — about 10%. Evidence you can handle both revolving credit (cards) and installment loans (car, mortgage, student). The lightest factor, and the one least worth engineering deliberately — it mostly takes care of itself as life happens.

The two readings that matter

Reading one: two behaviors are two-thirds of the score. Payment history (35%) plus amounts owed (30%) — paying on time and keeping balances low against limits — total 65%, and both are behaviors, not circumstances. You cannot quickly change your account ages or the applications already made; you can pay this month's bills on time and this month's balances down. The score is far more steerable, on far shorter timescales, than most people believe — a genuinely unusual property for a number this consequential.

Reading two: what's absent is as instructive as what's present. Income, assets, employer, education — none of it counts. Two consequences, both liberating in different directions: no raise or windfall will ever directly buy score points (only borrowing behavior moves the number), and no one is scored for being poor — only for how borrowing went. The score isn't a wealth grade or a character grade; it's a narrow prediction of repayment behavior. Treat it as exactly that, and the folklore (Credit Myths dismantles it) loses its grip.

The weights, translated to consequences

The practical translation: one missed payment — 35% territory — can cost roughly 60–100+ points from a good (700+) score, sometimes more, sometimes less depending on the starting point; the higher the score, the harder the fall.² Meanwhile the difference between 20% and 90% utilization — 30% territory — commonly spans dozens of points, and unlike the late payment, reverses immediately when the balances do (the no-memory rule, Credit Utilization Mechanics and Utilization Deep-Dive). The two behaviors within your monthly control govern the majority of your number — which means most of "credit repair" is just these two behaviors, done consistently, with the file watched for errors (Your Credit Report and Score's one-in-five error rate).

Income context, per the curriculum's rule: the score's financial consequences scale with borrowing — the half-point of mortgage rate it moves (Preparing Your Credit for a Mortgage) is worth tens of thousands to a homebuyer and nothing to someone who'll never borrow. But for thin-margin households the score is also the gatekeeper to escaping the extraction economy's worst tier — the difference between 36%-adjacent credit and mainstream credit (The Extraction Economy's predatory line) — which makes the two steerable behaviors disproportionately valuable exactly where money is tightest.

Where Plenee fits

Plenee doesn't show a score — deliberately. It tracks the two heavyweight inputs you actually steer: payment timing (via due dates, projected cash, and autopay-in-full — Late Fee Elimination's architecture) and every card's balance against its limit, per card, with statement cycles attached (Utilization Deep-Dive's dials). The score follows the behaviors; the behaviors are what a tool can genuinely help you run. Watch the inputs and the output takes care of itself.