Online reviews in numbers – the Trustpilot Transparency Report 2021

Recently, Trustpilot published, for the first time, its Transparency Report (here). This report concerns Trustpilot's activity in 2020. Trustpilot is an online platform where consumers and businesses can share feedback (and respond to it) through online reviews. The importance of online reviews in consumer decision-making is undeniable, and information on them allows us to better understand how and why consumers use reviews. From a consumer law perspective, it is important to assess how the information conveyed through online reviews impacts the conclusion of consumer contracts. The biggest concern in this regard is the proliferation of fake reviews, often written by businesses to harm a competitor’s reputation or to boost its own reputation. Perhaps most importantly, consumer law is (and should be) concerned with how online platforms design and manage online review mechanisms.

In this context, Trustpilot’s report provides a rare insight into the collection and processing of online reviews, given that most platforms do not reveal information on this. While most of what the report contains can also be found in Trustpilot’s guidelines on online reviews (available on their website), some interesting information is revealed. In 2020, Trustpilot received around 38,000,000 reviews. Trustpilot removed over fake 2,000,000 reviews, which corresponds to almost 6% of all reviews submitted in 2020. Most of the removed fake reviews were positive reviews (4- or 5-star reviews). On average, Trustpilot’s fraud-detection automated systems review 100,000 reviews per day. Additionally, in a clear effort to increase transparency, the report explains what factors are taken into account when calculating the score of reviewer satisfaction: time span, frequency, and Bayesian average. Remarkably, Trustpilot even justifies the choice of certain methods. For instance, Trustpilot claims it uses a Bayesian average to guarantee that businesses with fewer reviews are not unfairly punished for it. Besides, Trustpilot informs that, even though they have a freemium model in place, 92% of businesses who use its platform do it for free. Trustpilot refutes claims that it favors businesses that subscribe to the paying model, by revealing that paying businesses have an average review star rating of 4.39, while non-paying businesses have an almost equal average of 4.38. These numbers also show how the average star rating is quite high, given that the rating goes from 1-5. In fact, Trustpilot reveals that 71% of their 2020 reviews gave the highest score possible to a business. As said, insights like these – particularly originating from the online platform itself – are quite rare, so reading this interesting report is highly recommended. 

Dutch Authority for Consumers and Markets to take action against fake online reviews

At the end of last month, the Dutch Authority for Consumers and Markets (ACM) announced that it would be taking action against fake reviews, fake likes, and fake followers active on the Internet (here). The ACM identified these practices on platforms such as Facebook, Instagram, YouTube, and Google. This puts social media and influencer marketing – and associated practices perceived as harmful for consumers - in the spotlight. According to the ACM, these practices are harmful for consumers since they undermine the consumers’ confidence in the market, and they provide consumers with unreliable information. This is an important step in the regulation and enforcement of legal rules regarding online review mechanisms. It is not the first time that the ACM devotes attention to this problem. In 2017, the ACM published a study regarding online reviews mechanisms and suggested that their transparency should be increased to fight reliability concerns. The decision to take action against fake reviews and other misleading practices follows up on the guidelines on the protection of the online consumer developed by the ACM earlier this year (here). In these guidelines, the ACM explained that, among other aspects, a business cannot post fake reviews or instruct others to do so nor can it highlight only positive reviews or delete negative reviews. The ACM stated that it will reprimand businesses, influencers or other market participants that offer ‘misleading endorsements’ and, in case of refusal to stop their illegal activity, the ACM admits the possible imposition of fines.

Although online review-related commercial practices are covered by existing EU legislation (such as the E-Commerce Directive and the Unfair Commercial Practices Directive), the enforcement of such rules in this context is not frequent, although growing in the last couple of years. However, fake online reviews (or misleading online reviews that, not being necessarily fake, are biased due to, for example, social pressure, leading to a problem of reputation inflation) are a problem that affects many consumers. For example, a 2014 study conducted at the request of the European Commission (here) estimated that around 82% of consumers consult online reviews before purchasing a good. The EU consumer law regulation of online review mechanisms has been recently strengthened by the Omnibus Directive that, for the first time, explicitly imposed duties on traders regarding online reviews (see our blogpost on the Directive here). It will be interesting to see how Member States will transpose these measures and what framework will be applicable to online reviews at national level (e.g. competition law, contract law, advertising law).   

New rules on personalised pricing and fake reviews: EP and Council agree on the reform of consumer law

Two weeks ago we informed about the stage of the legislative process concerning the New Deal for Consumers, focusing on the proposal on better enforcement and modernisation of directives on unfair commercial practices (UCPD), unfair contract terms (UCTD), consumer rights (CRD) and price indication (PID). Earlier this week, the Council and the European Parliament reached a preliminary agreement on the proposal.

As we explained in the earlier post, separate positions of the EP and the Council on the commented file did not fundamentally differ. Short-term agreement on the proposal, therefore, does not come as a huge surprise (particularly in this heated period before European elections). Both positions rejected the original proposal of the Commission to impose limitations on consumers' right of withdrawal set out in the CRD. Aside from that, Council members did not particularly like further harmonisation of procedural rules, proposed by the Commission. Since respective proposals had not been particularly far-reaching in the first place, it was enough to further water them down a little bit to reach a compromise at the Council level (the situation is more complex with the second New Deal proposal, to be discussed in detail during the upcoming conference in Amsterdam). The EP did not fundamentally oppose.

On the enforcement side, besides restating that the level of penalties imposed on the trader should, where possible, consider penalties imposed for the same infringement in other Member States as well as clarifying the threshold of fines, which Member States should provide for in their legislation (at least 4% of the trader's annual turnover or 2 million EUR, if the information on turnover is not available), the agreed text essentially reflects the Council's position. More apparent deviations from it can be seen at the substantive level. Two particularly interesting points concern online reviews and personalised prices.

Online reviews

At the initiative of the EP, new provisions have been added to the UCPD to address the problem of fake online reviews. Specifically, the catalogue of information which should be considered as material under Article 7 has been extended to cover information "whether and how the trader ensures that the published reviews originate from consumers who have purchased or used the product". Consequently, if a trader provides access to consumer reviews of products and omits this kind of information, his practice might be qualified as a misleading omission.

On top of that, two additional points related to reviews have been included in the UCPD's black list. The list, set out in Annex I to the Directive, ennumerates commercial practices which are in all circumstances considered unfair (hence, without the need to verify other criteria, such as the impact on an average consumer). The new black-listed practices include:

23b. Stating that reviews of a product are submitted by consumers who have actually used or purchased the product without taking reasonable and proportionate steps to check that they originate from such consumers.

23c. Submitting or commissioning another legal or natural person to submit false consumer reviews or endorsements, or misrepresenting consumer reviews or social endorsements, in order to promote products.

All in all, the amendments on consumer reviews seem to be a step in the right direction. Article 7 essentially requires that information about potential origin of reviews is disclosed, while the black list lays down the consequences of disclosing it in a misleading way. Of course, one can wonder whether the black list is the right place for a provision whose substantive scope is not readily apparent. Thus, de lege ferenda, additional guidelines as to what "reasonable and appropriate steps" actually are would be advisable.

Black list cont'd

For the sake of comprehensiveness, two further amendments to the black list contained in the new text should be mentioned briefly.

Firstly, the compromise version features yet another new black-listed practice proposed by the EP, namely "reselling events tickets to consumers if the trader acquired them by using automated means to circumvent any imposed limit on the number of tickets that a person can buy or any other rules applicable to the purchase of tickets" (new point 23a). This seems to address a highly specific problem, which must have been brought to MEPs attention by respective stakeholder groups.

Secondly, the amendment about paid results in search engine rankings, which had already been proposed by the Commission, has been adjusted slightly. According to the current version (new point 11a) "providing search results in response to a consumer’s online search query without clearly disclosing any paid advertisement or payment specifically for achieving higher ranking of products within the search results" shall be considered unfair in all circumstances. At first glance this might seem like a step back from the earlier wording, which referred to direct and indirect payments. Note, however, that a reference to both types of payment can still be found in the preamble. 

Information duty on personalised pricing 

As regards the amendments to the CRD, a particularly interesting development concerns the scope of  traders' information duties. In this respect, a new point has been added to the list of information to be communicated to consumers in contracts concluded at a distance (Article 6), namely information "that the price was personalised on the basis of automated decision making" (where applicable, of course).

The origin of this amendment can be traced back to the IMCO report, yet its final scope does not fully reflect the original one. Firstly, the substantive scope of the duty has been narrowed down compared to EP's text. More specifically, the IMCO report referred to information "whether and how algorithms or automated decision making were used, to present offers or determine prices, including personalised pricing techniques". By contrast, the compromise text is limited to personalised pricing, and excludes other algorithmic techniques, such as dynamic pricing (recital 45). Furthermore, it only requires traders to disclose whether, and not how the price had been adjusted.

Secondly, and more favourably to consumers, in the compromise text the new rule is framed as one of general information duties for distance contracts, and not just a specific rule for contracts concluded via online marketplaces, as originally proposed by the EP. As a result, under new rules, consumers should at least get an overview of the application of personalised pricing across the digital market. This, of course, under the assumption that competent authorities will find a way of enforcing these rules.

Price information cont'd

The agreed text also includes another new provision about communication of prices, this time as regards price reductions. More specifically, under the newly introduced Article 6a of Price Indication Directive "any announcement of price reduction shall indicate the prior price applied by the trader for a determined period of time prior to the application of the price reduction". What is more, "prior price" is defined in the subsequent section as "the lowest price applied by the trader during a period of time which may not be shorter than one month prior to the application of the price reduction". Exceptions from this rule would only concern goods which are liable to deteriorate or expire rapidly and products which had been on the market for less than 30 days. How exactly this rule is to be interpreted, and especially how it would apply in the digital markets where prices can be changed dynamically, is a matter for a future reflection (on the application of the current rules, see e.g. our earlier post More transparency on hotel booking websites?). 

Final remarks on the level of harmonisation

Last but not least, several remarks about the level of harmonisation under amended rules are in place. The readers may recall that the original file proposed to change the level of harmonisation of Consumer Rights Directive from full to minimum in case of contracts concluded as a result of unsolicited visits at consumer's home and during excursions organised by a trader. This supposedly was a response to the calls from the Member States which sought to provide consumers in those circumstances with a higher level of protection. Ultimately, however, a different course of action has been chosen. Rather that changing the level of harmonisation, a few additional regulatory choices have been introduced to the CRD. Specifically, Member States would be able to extend the withdrawal period for these types of contracts from 14 to 30 days. Additionally, Member States would also be able to derogate from some of the exceptions from the right of withdrawal for these types of contracts. This, however, does not mean that the CRD in its entirely will now be based on the principle of full harmonisation. As readers may recall, already now Article 5(4) allows Member States to adopt or maintain additional pre-contractual information requirements for contracts other than distance or off-premises contracts. And finally, not to be missed, the newly introduced Article 6a(2) also introduces a minimum level of harmonisation with respect to disclosure duties for online marketplaces.