Wednesday, 6 October 2021

Trustpilot - Which? highlights areas that should be of concern to any business, and the regulator


On the face of the statement in the strapline - beginning 'In May...' - you would think a business would be crazy not to join Trustpilot, until you realise the organisation making the statement is Which? so it's obviously not meant as a compliment.

Let's mine a little deeper...

'In March, April and the first half of May, Sykes Holiday Cottages received more than 5,000 five-star ratings on Trustpilot. This was despite the fact that nobody could stay in one of its cottages at the time, because they were closed due to coronavirus.  At the same time, Which? was inundated with complaints about Sykes because of its policy of refusing refunds to customers for stays it had cancelled. There were so many that we reported it to the Competition and Markets Authority (CMA). Although it also received more than 2,000 one-star ratings in the same period, its flood of five stars meant that, at the time, it had an average of four stars and was officially ‘great’. 

It wasn’t the only travel company that faced widespread anger from customers over refunds. We found many heavily criticised firms rated as ‘great’ on Trustpilot when we looked at their profiles in early May. One example was online travel agent Kiwi.com – which has maintained a ‘great’ four-star rating despite customers complaining about difficulty getting refunds.  Yet others, such as British Airways, Expedia and Tui, which have also faced criticism, were rated as ‘poor’. Even the widely praised Hays Travel was rated as ‘poor’. 

And now the punchline...

'All but one of the travel companies we looked at which were rated as ‘great’ had paid Trustpilot hundreds of pounds a month. None of those rated as poor had paid it anything at all.

Here is Sykes' listing on Trustpilot...





A simple question for Trustpilot: how does a company that leaves 21 percent of its customers so dissatisfied that they feel obliged to write a one-star review rate as 'Excellent'? That's over one in five and ten thousand extremely unhappy customers (read some of the reviews if you doubt that). How many negative reviews does it take to get down to just plain 'Good'?

Now it starts to get even more sinister...

'How did Sykes manage to get so many thousands of people to leave positive reviews in such a short period? Sykes pays Trustpilot for an expensive ‘Enterprise Account.’ This allows it to use Trustpilot’s systems to invite thousands of reviews a month. In April it sent out 5,698 invitations and 60% of these received five stars. In the same period it received 718 reviews that weren’t from an invitation and just 20 of those were five stars. Trustpilot confirmed to us that, ‘the biggest driver of a higher score is whether or not a company invites its customers’. It said invites ‘tend to activate customers who had a good experience*, but who wouldn’t otherwise have shared their opinion’.  It also said that companies don’t have to pay to send out invites. They can ask customers directly to leave reviews, or use Trustpilot’s own free plan that allows them to send out 100 invites through its systems. But only one out of 62 companies that we looked at used the free plan. 

This statement goes exactly contrary to all our, and our clients', real-life experience: inviting reviews activates dissatisfied customers far and away above satisfied ones. That's a big part of why HelpHound exists: to moderate inaccurate and potentially misleading reviews pre-publication. To say the contrary goes so against our years of experience that we can only assume that Trustpilot's sales and marketing department had a hand in this incredibly misleading - and dangerous for any business considering adopting Trustpilot - comment.

While Sykes Holiday Cottages was rated as ‘great’ on Trustpilot, Hays Travel was rated as ‘poor’. Yet Hays was the third best holiday company in our most recent Which? Travel survey, with customers giving it five stars for service. It’s also been highly rated elsewhere for its response to the coronavirus pandemic.  At the beginning of May we looked at the ratings for 32 travel companies that pay Trustpilot and 30 that don’t. The average for those who don’t pay was less than 2.5 – officially ‘poor’.  The average score for those who pay was just under four stars – officially ‘great’. Even when we checked the scores again in July – after many months of almost no travel – the companies that paid Trustpilot still had a rating of 3.7, while those that don’t pay had 2.4.

'[Holiday] companies are also able to flag** – and potentially have removed – reviews that they don’t like. Kiwi.com flagged almost 800 negative reviews in the year up to the beginning of May. Trustpilot removed over 650 of them.  This helped it to retain its ‘great’ four-star rating on Trustpilot. Companies can dispute a review simply by claiming its ‘not based on a genuine experience’ – meaning a one-star reviewer can be asked to prove they used the company.  Yet five star reviewers are much less likely to be asked to prove that they had a genuine experience.  Kiwi.com told us: ‘When a review is received for which the customer can’t be verified we flag it to Trustpilot.’  Trustpilot said: ‘If any firm were to misuse the flagging function they would be breaching our guidelines and we would take appropriate action.’  

** We have addressed this 'flagging' facility more than once in previous articles: whilst superficially a 'good idea' it is almost universally used by businesses to place an extra obstacle in the way of customers posting negative comments (by insisting that they provide 'proof of purchase' to Trustpilot before their review is posted - can you imagine what the drop-off rate is?). How is that helpful to future customers of the business - who are relying on reviews to make an informed choice - that uses this facility in this way? It's not, of course; it's also against both the spirit and the word of the law

Here are Trustpilot's own numbers for another high-profile client: Purplebricks:



To be absolutely clear: PurpleBricks appealed against 427 reviews; all of those were suspended by Trustpilot (hidden from public view), Trustpilot then contacted the reviewers concerned and asked them to provide evidence of their dealings with PurpleBricks, 351 - that's all but 76 of the authors of the 427 reviews didn't respond to Trustpilot and further 19 failed to 'resolve the breach of Trustpilot's guidelines. Only 57 of those 427 reviews - a paltry 13 percent - ever saw the light of day. 

Our advice to consumers? Write your review direct to Google: it will be seen by so many more people and it will be published.

Our advice to businesses? Get reviews written to your own website and then to Google. Crucially, employ an independent moderator to ensure your reviews contain the minimum of factual errors and statements likely to mislead consumers.


Here's Hays Travel's Trustpilot listing...




On the surface, and according to Trustpilot: a 'Poor' business. But really? What we have here, in reality, is a business that refuses to engage with Trustpilot. Why? Not because it sees the obvious flaws in Trustpilot's offering, but because it uses a rival system over which hang perhaps even more question marks...





Hays have simply opted for Trustpilot-lite - or Feefo as it is properly known. Now with Feefo the business not only controls who gets invited to write a review - invitees of the business by email and absolutely no-one else - but it also controls the timing of the invitation to write the review (time to read the CMA regulations - link above - again!).

Now we have two methods by which businesses might, just might, be able to manipulate the numbers in their favour...

  1. Trustpilot: don't invite customers you know to be unhappy, but run the risk that they'll find their way to Trustpilot anyway; although once they have written an uncomplimentary review there you can make use of Trustpilot's quarantine facility, whereby you may challenge the reviewer to provide proof of purchase otherwise their review goes into limbo, invisible to consumers.
  2. Feefo: don't invite customers you know to be unhappy in the secure knowledge that they cannot write a review to the platform in any other way.
We suppose it could be worse: at least Yelp, the largest quoted review site on the planet, has scurried back to the States and is no longer offering paid listing uplifts to UK businesses. But it is past time for the CMA to crack down on both the businesses and the mechanisms described here and in Which?'s article (something Which? evidently agrees with - see below).

It's only fair that we reproduce Trustpilot's response to Which? although even we were surprised just how weak it was...

'Trustpilot told us [Which?] that although inviting customers to leave reviews is the biggest driver of a higher score***, the process of inviting does not guarantee a higher score, nor do companies have to pay to invite customers. It said that ‘on every company profile page, we make it clear whether the company is actively inviting reviews or not – but the causation between paying and high scores does not exist’. It has also pledged, ‘to fight for and maintain trust in online reviews’ with several new initiatives. These include allowing ‘flagged’ reviews to remain online while they’re being investigated, no longer sharing reviews with search engines for companies that have breached its guidelines, improving transparency and banning incentives to leave a review. This week it announced the launch of a new research and development hub in Edinburgh, ‘to develop new, world-leading technology that proactively tackles the behaviour that threatens trust online.’ 

Action from the Competition and Markets Authority? The Competition and Markets Authority (CMA) recently announced the second phase of its investigation into online reviews on major websites. Which? wants the regulator to investigate how sites collect, display and moderate reviews and whether that risks misleading people. It should also look at the robustness and transparency of sites’ T&Cs and how effectively these are enforced. If the CMA finds that sites are failing to take sufficient measures to protect consumers, Which? expects the regulator to take strong enforcement action.


*** Do we need to say 'No, it's not' again?

And here are links to articles relating to CMA regulations and action regarding reviews:

When you look at some of these dates you will see why we have become mildly frustrated by the CMA's relative inactivity. We have written to them on behalf of consumers in general and our clients in particular because our clients - who all obey both the spirit and the letter of the law - continue to find themselves competing on a playing field that is unfairly tilted against them as a result of the abuses by clients of sites such as Trustpilot and, to a lesser extent, Feefo, highlighted here.

It is testimony to both our clients and HelpHound, and to consumers in general, that this disadvantage has not translated into a lack of success in the marketplace. Just see how one prominent Plc client of ours has won market share by compliantly engaging with online reviews.

No comments:

Post a Comment

HelpHound is all about feedback, so please feel free to comment here...