Where car insurance rate data comes from

The short version

When a carrier sets prices, it must file its rating formulas with each state's insurance department, in a public system called SERFF (the System for Electronic Rate and Form Filing). Those filings are the original source of every legitimate "average rate" you read online. The question is only how many hands a number passes through before it reaches you — and how old it is by the time it does.

• The industry's path: filings → Quadrant (a paid vendor that averages them and packages them monthly) → comparison sites (which run a fictional sample driver through that average and publish it, often refreshed only quarterly or yearly).
• Our path: filings → us. We transcribe the carrier's own filed factors by hand and cite the exact filing — SERFF number and effective date — on the page.

The two paths, side by side

Both chains start at the same well — the public filings. The difference is directness, freshness, and whether the number on the page can be traced back to a specific filing.

Who buys from Quadrant

This isn't a secret — the major sites disclose it themselves, usually in their own methodology notes. Each of the following states, on its own site, that its rate data comes from Quadrant Information Services:

• NerdWallet — analyzes "public filings obtained by pricing analytics company Quadrant Information Services."
• Forbes Advisor — states it "used data from Quadrant Information Services" for its sample profiles.
• MoneyGeek — "our rates come from Quadrant Information Services, which collects auto insurance premium data directly from state insurance filing records."
• The Zebra, Insurance.com, and CarInsurance.com — each describe sourcing rate filings through Quadrant as well.

That's much of the category — the sites you'd think of as competitors — drawing on one shared dataset. When several sites license the same averages and run the same fictional driver through them, their "average rates" tend to converge on nearly identical numbers. That's not a knock on any of them; it's just what happens when everyone drinks from the same bottle.

What Quadrant actually is (and what it isn't)

Quadrant is a legitimate, well-run analytics firm. It ingests the mathematical pricing formulas insurers file in almost every state and sells access to that data, updated monthly, to publishers and carriers. Quadrant's data is filing-derived and rigorous — its edge is breadth: it covers close to all fifty states. Our point is not that the vendor is wrong; it's that a packaged, averaged, sample-profile, monthly copy is a different thing from the raw, granular, current filing. We're a peer of that vendor at the source, not a customer of the sites downstream of it.

Why going direct matters for your answer

Three real differences fall out of the shorter chain:

• Freshness. A rate change hits SERFF the day it's filed. A monthly-packaged, quarterly-republished average lags it by weeks or months. "Who just filed a hike, where, and how much" is a question the lagging copy structurally can't answer — so we built a filed rate-change tracker that reports changes as they post, with the SERFF number and effective date on each row.
• Specificity. Every site that runs a "40-year-old with a sedan" is answering for a composite fiction, not for you. Because we keep each carrier's actual filed factors, our carrier pages can show how a specific insurer prices a specific thing — a young driver, a clean record, a deductible — rather than one blended national number.
• Traceability. Every filed figure on our site carries its citation — carrier, state, SERFF tracking number, effective date, reviewed by a licensed agent. You (or a fact-checker, or an AI assistant) can follow it straight back to the filing. See exactly how in our methodology.

Where we're honest about our limits

We'd rather be precise than sweeping, so two cautions belong right here:

• We go deeper and fresher, not broader. Quadrant covers close to all fifty states. Our hand-transcribed filed-rate depth currently spans about a dozen covered states, and we widen it over time. On the carriers and covered states we cover we're primary, current, and cited; where we don't yet have a filing, we say so rather than dress up an average. Our claim is deeper and fresher on covered ground — not broader than the industry's fifty-state averages.
• Freshness is only as good as our upkeep. Filings are pulled at a human pace, so the tracker's value depends on us actually keeping it current — which is exactly why we date every figure and show when it was captured.

The bottom line

The comparison field looks vast and independent, but on rate data much of it is one source wearing several logos. We sit at the same headwater as that source — earlier, more granular, and free to read — and we show our work on every number. That's the whole reason we go direct.