We do not compare live prices. We compare estimates, and we show you how they are made.

Every comparison site in Greece shows you a number and never tells you where it came from. This page is the other option: here is the whole model, together with the places where we know it is wrong.

The basic idea: a reference driver and multipliers

Insurance companies do not hold a ready price for every case. They have a base premium and a series of factors that move it up or down. That is how they price internally, and that is how we do it too.

Your own price comes from multiplying each company's base premium by one factor for each thing that sets you apart from that driver. That is exactly what “How it was worked out” shows under every row of the comparison: not a summary of the calculation, but the calculation.

The factors, with real numbers

Every factor of the model, for all 28 products, is published live on the page of each product, at the foot of this one: the same steps for every company, different heights, and a control under each figure to see what one change does to the price. Geography stays here, because it is where the model and the claims data say different things.

Area

Five zones and not thirteen regions. The administrative map does not price: Patras and a village in Achaea are the same region and not the same risk, while Patras and Heraklion are different regions and are priced almost the same.

The last two columns are not prices - they are what the claims cost. They come from the ΕΑΕΕ's Motor Insurance Statistical Yearbook 2024, which pools 24 companies - about 93% of the market - and reports, for every prefecture, how often a third-party liability claim happens and what one costs. The two multiplied together are the euro per insured vehicle per year.

The per-zone figures are the mean of the prefectures in each zone, weighted by population: ΕΑΕΕ publishes vehicle counts by vehicle use but not by prefecture, so the resident population of the 2021 census stands in for a vehicle count - a proxy, not a measurement - and Samos no longer counts as much as Attica inside its zone. The error that leaves is still measurable - the five zones weighted by their share of the fleet land -6.8% from the national figure ΕΑΕΕ publishes itself (6.87% × €1410).

Where the numbers come from

Every price on this site is our own estimate. It is not an offer and it is not a price from a company. We hold no rate book from any insurer, we connect to none of them, and none of them sends us prices. Every row of the comparison comes out of the calculation this page describes.

The public data. The calculation rests on figures somebody else publishes and you can check yourself, chiefly: ΕΑΕΕ, the Bank of Greece, ΕΛΣΤΑΤ and Eurostat, the annual solvency reports of the companies themselves, and the published terms and information documents of the products. Every figure we take from them appears with its source and its data year, on the page that uses it.

Our own assumptions. The shape of the curves, that is how far the price climbs with age, with engine size or with region, is published by nobody in Greece. It is our assumption, we write it out in full on this page, and it is where the calculation can be most wrong. Every estimate is a gross premium, with the insurance tax and levies included.

The calculation can be wrong. It does not necessarily reflect the current state of the market: companies change their pricing without announcing it, and the price they give you can come out above or below our estimate. The order of the ranking is more reliable than the amount itself. Only the company gives a binding price, after an application.

Where it goes wrong

Four things, and we say them because you will find them out anyway.

• We do not ask everything the company asks. Their form has about thirty fields. We ask seven, the ones that really move the price. The rest (make, model, kilometres a year, whether it sleeps in a garage) move the premium a little, and that “little” is part of the range we show beside every price.
• We do not see the offers. Companies run discounts: summer ones, for new customers, for two vehicles, through a particular agent. None of them is in the model. In practice this means the price the company gives you comes out more often below our estimate than above it.
• The claims history is your word. There is no public bonus-malus register in Greece that a site like this could read. If you say “never a claim” and the company sees otherwise, the price it gives you will be noticeably higher than ours.
• Some companies simply will not insure you. A driver under 23 with a one-year licence and a 2,000 cc car: several companies do not write such a policy at all. We will show you a price. The company may say no.

How often it is updated

We hold no company prices, so there is no old price here that can pass for a fresh one. What ages is the public data the calculation stands on, and all of it carries a date.

• Every source appears with the year of its data, beside the figure it supports.
• On every build, the check counts and states how many rows are estimates. Today it is all of them.
• On every build, the reference premiums are checked against the latest ΕΑΕΕ yearbook. If they drift, the build fails and the page is not published.

The last check is deliberately harsh. An old figure looks exactly like a fresh one, and nothing in the look of a number shows how old it is. So we do not leave it to our own judgement: the check stops the publication by itself.

What we do not do

We do not sell insurance, we are not an agent, we take no commission from any of the 20 companies in the comparison, and we do not ask for contact details to show you a price. There is no “send me an offer” here, because there is nobody at the other end to send it.

The reason is practical, not moral: the moment we took commission from one company, the ranking would stop being a ranking. You would not know by looking at it; nobody knows by looking at a ranking, and that is why the only answer that means anything is for the commission not to exist at all.