Let's be honest about what Trustpilot reviews tell us, and what they don't

Last night while I was doing the washing up - an advert for LV= came on the radio. I can't remember all of the boasts that it made - but as with most insurance advertising, it was naturally all about how they'll be there for you when you need them. As I’ve written before – I do worry that insurance advertising plays a part in widening the gap between what customers expect and what they get. But that’s not what this blog is about.

What really caught my attention was the last line of the ad - "Perhaps that's why we're rated excellent on Trustpilot". But it’s not.

If you don’t know how Trustpilot’s model works – it’s important to understand it. Trustpilot can certainly be useful in helping you understand how good a firm is – but the headline ratings are rarely the best indication.

Firms that get a good score on Trustpilot are almost universally firms that pay Trustpilot to poll their customers. You can always spot a firm that doesn't have a commercial relationship with Trustpilot because they have a terrible rating. Their Trustpilot page just becomes a repository for grumpy customers to vent their spleen. The best example of this is in banking - where Nationwide gets an overall score of 2 stars. Nationwide recently emerged as the most trusted banking brand bar none in our polling - beating even the digital challengers like Starling and Monzo. But it doesn't have a relationship with Trustpilot, so its scores on their site are poor.

Confirmation bias

Trustpilot works by becoming your polling partner – helping you solicit reviews from your customers. And surveys tend to go out just after you’ve made a purchase. You may even be asked to complete a survey directly after the end of your purchase. This creates a confirmation bias. People who have just bought something tend to say that they’re happy with it. But they are not in a position at that stage to tell you what the firm is like over the long term.

In insurance, these ratings are a little more divisive than in some other sectors, where an instant view from a new customer might be useful. If you’ve just bought a pair of trainers, you can at least judge the firm based on the quality of the product when it arrived, or the speed of delivery and the customer service.

But in insurance, customers who have just bought a policy know nothing about the most important part of their new purchase: will the insurer be any good if they have to make a claim?

There are some insurers who get good reviews on Trustpilot but who I know deliver a fairly sub-optimal experience for their customers at claims stage.

This is a genuine problem in regulated markets. These reviews give consumers a false sense of security, and lend credibility to brands who may not deserve it.

Busting myths

I can’t tell you how many times over the years that I have presented our polling results to companies, and have been told that these can’t be true because they have a great score on Trustpilot.

Recently, a member of senior staff in a trade body told me that their members’ great scores on Trustpilot were proof they were all great.

But just because the numbers are telling you what you want to hear – it doesn’t mean they’re right.

This is not designed to be an out and out attack on Trustpilot. As I said at the top, I think it does have validity as a service. But personally, I always find most value in looking at the one and two star reviews – as these are where you find out what’s really going on behind the scenes.

I think Trustpilot is well aware of this problem with its business model – but chooses to look past it. I’d like them to admit it – and think about how they could work to create something that really does help consumers understand the quality of insurers.

We need real competition on quality

As things stand, there’s no reliable way to understand which insurers will keep their promises and which drag their heels and prioritise cost control over good customer outcomes.

We do our best to get an answer through our customer experience ratings – which add Ombudsman uphold rates and our own transparency analysis into the mix. But I would be the first to admit that this is not perfect – and of course we don’t get a big enough sample for many of the smaller brands in the market.

I’d like to see the FCA create a standardised way for insurance claimants to be polled – with firms mandated to publish the results. If we’re going to have competition that works in the insurance market, consumers need to have reliable data to work with.

In the meantime, firms need to stop pretending that Trustpilot reviews prove something that they don’t. It’s misleading and puts them in breach of FCA rules.