Misleading press release from Go.Compare?

Earlier this month, Go.Compare issued a press release, "Driving a bargain: Car owners could save £239 by paying annually for insurance, new data shows".

As you will no doubt agree, the title of this press release communicates that a motorist who opts for an annual payment instead of monthly payments could save £239 on average. The rest of the press release (below) says more of the same. For example:

"According to the insurance experts at Go.Compare, the average cost of a car insurance policy when paid annually is £418, compared to £657 when paid monthly - a 57% increase" and "depending on whether you choose to pay monthly or annually... average saving of £239 for an annual policy."

In NimbleFins' view, this conclusion is quite misleading for consumers. Go.Compare implies that for a given motorist, the difference between monthly and annual payments is £239 or 57%. Yes, we agree that you can achieve savings by paying annually, but we think it is simply not true that, on average, a given motorist would pay anywhere near ~57% more by paying monthly rather than annually. And yes, insurers may charge up to 30-40% or more in certain cases for "risky" motorists, but how many people actually pay 57% extra simply for paying their car insurance monthly (that would make a good study!)?

Here's what we think went wrong with the GoCo data analysis...

First clue: the common sense market check

We've spent years looking at car insurance quotes, and think a common price difference between paying annually and monthly for a given policy is ballpark 10% (or a bit more) for a lot of motorists (not 57%). Yes, some people will pay more (e.g. could someone using a broker and essentially getting a loan from a lender for car insurance pay interest rates in the high teens or 20s? Sure. Could a risky driver be charged 30% or even 40% extra? Sure.) But for the bulk of policies, from the insurers underwriting the most policies, the added cost would be as low at 9-10%, maybe up to 20%. Not 57%.

To check ourselves, we ran some test quotes for a sample driver of a Ford Puma (because it's the most popular car in the UK) on a leading car insurance comparison site. Here were the five cheapest quotes returned, showing both the annual and monthly payment options. (Note, we picked a profile that has a car insurance cost very close to the average premium of £418 mentioned by Go.Compare, so we can see if our sample driver would pay a similar bump up of £239 for paying monthly—they don't.)

Note, the quotes above reflect leading insurers like Allianz and LV=.

As you can see, the average premium was £429 for those paying upfront and the average pay-monthly total cost was £472, a difference of £43 (£472 - £429 = £43) or 10% extra for paying monthly. Nowhere near £239 or 57% extra that Go.Compare claims it'll cost you for paying monthly.

We also ran test quotes for a "risky" driver (young and unemployed)—because it makes sense that a risky driver with a higher premium might be charged a higher borrowing rate. The result? This driver was being quoted policies starting from around £2,000 per year. And not many insurers even quoted so not a lot of options; and sometimes insurers wouldn't even offer monthly payments. When monthly payments were offered, we saw additional costs of up to 30-40%. So, yes, it can cost quite a bit extra to pay monthly. But not 57% on average across all motorists.

So, we have confirmed in the marketplace that the additional cost of paying monthly for a car insurance policy is well shy of 57% extra, as we suspected. So why has Go.Compare come up with such a strange conclusion and press release?

Second clue: the data check

Go.Compare is comparing apples with oranges. They think they're comparing the additional cost to a motorist of paying monthly vs. annually, but they're not. They're comparing two different cohorts of motorists (those who opt to pay monthly vs. those who opt to pay annually).

To recap, they said that the average cost of car insurance when paid annually is £418 (we have no dispute with this piece of data on its own), and the average cost of car insurance when paid monthly is £657 (we have no dispute with this piece of data on its own), and concluded therefore that the average difference to a motorist between paying annually and monthly is £239 or 57% (here's where we take issue).

Go.Compare is drawing a conclusion ("paying monthly costs an extra £239 or 57% on average") as if comparing two otherwise-equal baskets of policies—that only differ by payment method (monthly vs. annual).

But Go.Compare is comparing two baskets of policies that surely differ by a lot more than just payment method. We suspect that the basket of pay-monthly policies is full of policies that are naturally higher-priced, whether due to driver's age, history, car, or other factors that impacts car insurance premiums.

It makes sense—if we can generalise for a moment, people with high car insurance premiums are more likely to need or want to spread out the high cost by paying monthly; people with low car insurance premiums will have an easier time paying the whole premium upfront. Who wouldn't be more comfortable paying a £400 premium upfront than a £1,200 premium? That's the whole point of the monthly payment option.

So we can't say for sure without digging into the data ourselves, but we suspect the data would show a simple consumer behaviour: the higher the car insurance premium, the more likely someone is to pay monthly. (That would also be a good data study!)

Anyway, all of this means that Go.Compare's basket of pay-monthly policies is probably made up of policies with higher underlying rates. And that's why Go.Compare is seeing such a high average cost of car insurance when paid monthly of £657. Not just because of an excessive additional cost for just paying monthly—but because those who pay monthly naturally tend to have higher car insurance rates.

What could you say from the Go.Compare data?

One conclusion you can safely draw from the Go.Compare data is that:

Of this 57%, if only around 10 percentage points (e.g. ~£60 on a £657 policy) is likely due to paying monthly instead of upfront (i.e. the borrowing costs from spreading out the upfront cost over the course of a year), then the other 47 percentage points or so (e.g. ~£180 of the £239 from the GoCo data) would be down to other factors that impact the cost of car insurance, like the driver's age, history and car. For a driver being charged 20% or even 30% for the monthly payments, much of the monthly-vs.-annual difference is still likely due to these other factors, not simply the choice to pay monthly.

A second conclusion you can probably draw from the Go.Compare data is that, on average:

In reality, if the pay-monthly additional cost is around 10%, this means the typical pay-annually motorist would pay around £42 extra (10% of the average annual cost of £418) to switch to monthly payments; if the pay-monthly additional cost is around 20%, they'd pay around £84 extra.

It could be that monthly payers have even more to save—if these motorists do have higher rates (which we think), they may be riskier and face higher borrowing costs. If a typical pay-monthly motorist (whose insurance costs £657 in total for all of the payments) does so at a cost of 20%, they could save around £110 on average by switching to an upfront, annual payment. But we're hard pressed to think of a scenario where motorists on aggregate pay 57% extra for monthly payments.

We aren't privy to their data but, based on our experience, Go.Compare should NOT claim or imply that a motorist who opts for an annual payment instead of monthly payments would save £239 on average or that it costs 57% extra to get monthly payments. We think that figure is based on incorrect data analysis and misleading for consumers by way overestimating the cost differential between annual and monthly payments for a given motorist—by a factor of potentially up to 4X - 6X.

A better way to dig into the data

A better study could be, "Who pays their car insurance monthly?"—sort the data into premium buckets and compare histograms for those paying annually and monthly. You could see, for example, what proportion of people with premiums in the £300 - £400 range pay monthly vs. annually, and so on. Or look into different cohorts like preferred payment option by age bracket, car value, etc. That could be both interesting and useful.

We have reached out to Go.Compare for comment and will update this article as needed with their response or any insights we may have missed.

Press Release

Go.Compare's Press release: "Driving a bargain: Car owners could save £239 by paying annually for insurance, new data shows":