How we compare car insurance companies

Each comparison page puts two carriers side by side on published, third-party figures: state-regulator complaint data, financial-strength grades, coverage lineups, app ratings, and a set of product facts (SR-22 filing, telematics programs, who underwrites the policies). Every figure carries its source, the entity measured, and the year, printed under its row.

We do not publish composite scores or star ratings. A blended score depends on weights someone chose, and the weights that matter are the reader's. We publish the inputs and state, per row, which side the figure favors.

Complaint figures come from the National Association of Insurance Commissioners' Consumer Insurance Search complaint index. An index of 1.00 means a company drew exactly as many state-department complaints as an insurer its size is expected to; 0.50 means half as many, 2.00 twice as many.

The NAIC publishes this index per underwriting company, not per brand. Large groups write through dozens of entities, so we name the entity we cite and use the group's largest private-passenger auto writer by premium. We also show the raw complaint count and premium volume, because an index built on four complaints means less than one built on four hundred.

Some sites blend the index across a group's subsidiaries. That is a defensible method; we prefer the single named entity because it can be checked directly against the NAIC's own report. Complaint figures on our pages compare carriers only within the same reporting year, and only on private passenger auto.

One caution we repeat where it applies: non-standard auto policies involve more lapses, cancellations, and payment events than standard ones, and each is a chance for a dispute. A non-standard writer's complaint index tends to run higher than a standard carrier's for reasons that are partly about its segment, not only its service.

Financial-strength grades come from AM Best. We show the letter grade and its category. Where AM Best has not rated a carrier, the row says so or does not render; we never substitute another agency's grade and present it as comparable.

App figures come from Apple's App Store, with the rating count and the month we read them. We publish a rating only when the app has at least 1,000 ratings; below that the figure is noise. Ratings within 0.2 stars are called a tie.

Car insurance rates are filed state by state and priced to the individual driver. The same violation can move two carriers' quotes in opposite directions, and non-standard filings vary between states far more than standard ones. A single national price comparison would be a guess presented as a fact, so the price row on every page is unscored and the pages state what actually sets the number: the driver's record, the state's filings, and the difference between money due at signing and the term total.

Truvo is a licensed independent insurance brokerage. We quote the carriers we can place in a reader's state, standard and non-standard, and place drivers with whichever comes back best.

Comparison figures live in one content system, so a corrected or refreshed figure updates every page that shows it, and each page's updated date reflects the latest edit to either carrier's data. Complaint indexes update when the NAIC publishes a new reporting year; app ratings are re-read periodically with the read date shown.