In late 2018 my Experian FICO was 580. I had two credit cards maxed out from a six-month period of unemployment, a 90-day-late mark on one of them, and a small medical bill in collections that I had ignored. By spring 2020 my score was 760. The rebuild was less dramatic than people assume — there was no clever trick, no credit-repair company, no secret-shopper credit card. Mostly it was a four-step plan that I executed for eighteen months and then stopped thinking about.
How I broke it (briefly)
In April 2018 I was laid off. I had three months of savings, which became six months of unemployment. By month four the savings were gone and my fixed costs were going on credit cards. By month five I missed a payment on one of them — not because I could not pay, but because the autopay had been linked to a checking account with insufficient funds. The 90-day-late mark, mechanically, dropped the score harder than any of the maxing-out.
I am writing this six years later in case there is anyone reading this who has done a version of the same thing. The score was the lowest it had ever been and the most embarrassing thing in my financial life at the time. It is, in 2026, a footnote.
What I did in the first month
Three things, none clever:
- Set up autopay for at least the minimum on every account. From a different checking account this time. Late payments are the single largest weight in the FICO model. A second late payment would have set the rebuild back by months.
- Paid the medical collection in full. $640. I called the collector, asked for a "pay for delete" — they remove the account from my report in exchange for full payment — and they agreed. About 50% of collectors will do this if you ask politely. It is not a guarantee, but it costs nothing to try.
- Pulled all three reports. Free, via AnnualCreditReport.com. Found two errors. Disputed both. One was corrected.
The four levers that actually moved the score
Payment history (no further misses)
Eighteen months of clean payments was the biggest single factor. The 90-day-late from 2018 stayed on the report for seven years, but its weight in the score model drops as more months of on-time payments accumulate behind it. By month nine of clean payments, the late mark was costing me maybe 30 points; by month eighteen, closer to 10.
Utilization (got below 10%)
This was the fast lever. Credit utilization — the share of your available credit you are using — is recalculated every time your card issuer reports a balance, typically once a month. Getting from 92% utilization to 7% added about 60 points in two months. The mechanics: I paid down the two cards aggressively for nine months until balances were under 10% of limit, and then asked both issuers for credit-limit increases. Increases were granted. Utilization fell again.
Average account age (kept old cards open)
Counterintuitively, I did not close the older of the two cards even when I had paid it off and stopped using it. Closing it would have shortened my average account age, which is roughly 15% of a FICO score. I let it sit, used it for a Netflix subscription, and paid it in full each month.
Credit mix (added one installment loan)
Small effect, but real. About a year into the rebuild I financed a used car at a credit union. The installment loan added a second tradeline type to a file that had been credit-cards-only. Worth maybe 10 points over the following twelve months.
Credit scoring rewards boring people who have been boring for a long time. The rebuild is essentially the slow accumulation of evidence that you are now boring.
The things that did not matter
- Credit-builder loans. I tried one for three months early in the rebuild. Tiny positive effect on the score, real monthly cost. Not worth the friction once I had the basics in place.
- Secured cards. I had unsecured cards still open, so secured was unnecessary. They are useful if you have no open tradelines; not otherwise.
- "Credit repair" companies. The CFPB has been clear for years that anything a credit-repair company can legally do, you can do yourself for free in an afternoon. I did not pay one. Nothing they would have done would have helped.
- Adding myself as an authorized user on a family member's card. I considered this and decided not to bother. The score bump is unpredictable and reversible.
What I would tell past-me
The lowest score is not a permanent record. The mark from 2018 will fall off in 2025; the score will keep moving regardless. Treat the rebuild as a slow project — autopay set, balances down, files audited once a year — and stop refreshing the FICO simulator. It moves on its own schedule, mostly without you.
