Car Insurance: Comparing the Comparisons…

As regulars will know, one of the key steps in the Cheap Car Insurance guide, is to use comparison sites in the right order. To decide that order we undertake a comparison of the comparison sites. The sheer scale of MSE and the fact the papers often cover the order, means it’s hotly contested.

Recently we’ve seen a little more lobbying and occasional calls too far asking things along the lines of ‘we’ve added Meanothing Insurance, can you change our numbers and move us up the ranking?’. So to help communicate the system better and explain exactly what we’re doing, I’ve just sent out a big e-mail to all the sites, which I also thought’d be useful to blog.


Dear folks,

The questions have been flowing through to my team and I about the order in our car insurance tables, so I wanted to clear things up. Hopefully it’ll help you help us to get even better comparisons for our users and make all our lives easier.

The logic.

No one comparison site covers the whole of the market, even in combination there are still some missed, yet my aim is to ensure that in the shortest amount of time, by combining sites people can cover the widest range, with accuracy, ease and no hassle. Our order reflects that and it is carefully put together.

How often it’s updated.

I know there’s been confusion on this so I wanted to notify you of the new system. We will do a full survey roughly every three months, and this will then hold firm unless there is a seismic change (or we are doing a full article update) and it needs to be done more radically.

Yet in general small changes to numbers of providers included will NOT be updated during that period. Yet so that our checks can remain random rather than expected, the exact period of the survey will move slightly.

This three-monthly system will be communicated to our users, so they have clear expectation of the survey period. However the section covering specific providers deals (lower down, for cashback offers and things) will be updated weekly.

How we analyse the order.

This is based on a number of factors….

  • Number of different insurers/brokers included. This is the lead data, but is only ever one of the factors. However the order is more specifically about “additional insurers” that you add on top of what’s already been used before you.
  • Number of quotes returned. We also separately do our own random independent checks on how many actual different competitive quotes are returned for a number of profiles. You will not be told this is happening, this is done as normal users. So a comparison site listing lots of providers but only actually giving results from a few will be downgraded.
  • Feedback. We receive huge amounts of feedback both by email and on the forum on providers, and we use these to help (though our own judgment is also used) us assess ease of use, accuracy and overall performance. This does leave newer comparison sites at a disadvantage to older ones, as there’s less feedback, but we consider in such a competitive market “tried and tested” has its advantages, and a good new site will be able to prove itself quickly and soon move up.
  • NEW. Follow up phone calls. We’ve had a lot of complaints about these, so wanted to give you advance notice that from our next review we will be including this as a major part of the process. Firstly we will note whether companies offer an ‘opt out’ or ‘opt in’ to follow up phone calls option – the preference obviously being an opt in system. We will also consider the actual feedback of how many phone calls people receive. Quite simply after doing a comparison, people get pee’d off with five or six follow up calls from different insurers.

Of course, some of these factors could be manipulated. For example we know a number of insurers provide identical quotes as they use the same systems, so it’d be easy to deliberately add these to comparisons to bulk up the numbers. To put your minds at rest, don’t worry, if we see any one doing this (or similar tricks), we’ll drop’em like a stone. In fact so much so, for a bit of fun, I’ve asked Richard (one of the MSE team who used to work on the other side – ML) to come up with the tricks he’d play if he wanted to cheat, so we’ve a full list of things to watch for.

As I say, the aim of this note is to give you a legitimate expectation of what we’re looking for, in the hope it will make life easier for all of us, and provide an even better result for users.

Kind regards,



Comment and Discuss