Digital Markets, Competition and Consumers Bill

Written evidence submitted by Trustpilot (DMCCB09)

Committee Stage Written Evidence

The Digital Markets, Competition and Consumers Bill


1. Trustpilot welcomes the introduction of the Digital Markets, Competition and Consumers Bill. As an online reviews company which has a mission to build trust online and which prioritises safeguarding our platform from bad actors, we strongly support the Government’s aims of ‘protecting the integrity of the review system’ [1] and ‘improving the authenticity of online reviews’ [2] .

2. Online reviews have become embedded in the retail experience. They are a valued source of information for consumers ("It is an invaluable tool for getting genuine opinions from other customers" [3] ), helping them to make informed decisions ("It's a fantastic reference point before deciding whether to purchase from a company" [4] ), and also for businesses, enabling them to gather feedback from their customers, refine their offerings and grow their customer base. In turn, this helps to stimulate competition as consumers can voice their views and discover new businesses, whilst businesses can hone their products and services, and attract new customers. [5] In respect to this latter point, review platforms can be used as a source of "free" promotion for businesses, as online reviews surface and amplify word of mouth from their customers to help prospective customers.

3. Against this backdrop, some bad actors have sought to take advantage through the sale and use of fake reviews. It is therefore positive that the Government is seeking to address this issue.

4. This submission sets out Trustpilot’s response to the approach for fake reviews being taken in Part 4 of the Digital Markets, Competition and Consumers Bill. Overall, we think the legislative approach is the right one, however it could be strengthened by including sites which host the sale of fake reviews within its scope. Moreover, it is critical that a pro-competition approach is taken to the world of online reviews. When the Government’s proposals for fake reviews are fleshed out they must take account of the range of business models and types of reviews in the landscape and avoid a one-size-fits-all approach which could stifle the consumer voice, competition and innovation.

The legislative approach

5. The approach put forward in part 4 of the Bill to secure delegated powers through which the Government will amend the Consumer Protection from Unfair Trading Regulations (CPRs) to address fake reviews is a sensible mechanism.

6. In Annex 4 of the Bill, the Government details that it plans to use the delegated powers from the Bill to add to the list of banned practices within the CPRs so as to address fake reviews. The proposal is to ban:

i. the commissioning or incentivising any person to write and/or submit a fake consumer review of goods or services;

ii. hosting consumer reviews without taking reasonable and proportionate steps to check they are genuine; and

iii. offering or advertising to submit, commission or facilitate fake reviews.

7. The Government intends to further consult on the details of this proposal. Trustpilot welcomes this initial approach, however there are a number of key factors to consider, which we detail in the following sections.

Addressing sites hosting the sale of fake reviews

8. In order to successfully address the issue of fake reviews, it is important that all the stages of the process are included. As it stands, the Government proposes to address those selling fake reviews (point i), platforms hosting reviews (point ii) and offering or advertising to submit fake review services (point iii).

9. Currently the proposed approach looks at the beginning of the fake reviews process with those selling fake reviews - tackling the issues at source via point i and iii - and the end, with the platforms which host reviews (via point ii). It is however missing what happens in the middle, namely the ISP and social media sites which host the sale of fake reviews. It is widely known and evidenced that review sellers use private groups on social media sites to advertise the sale of fake reviews. Indeed, the Government’s fake reviews research drew on a dataset of fake reviews collected from private Facebook groups where sellers buy reviews. [6] As such, it is vital that onus is also put on these sites to ensure that they are responsive to requests made by other parties who identify and request the removal of fake review sale groups on their sites, alongside taking proactive steps themselves to identify and shut down such groups.

10. To this end, we would advocate that a fourth area be added to the list of banned practices under the CPRs to include the element of ‘hosting the offer or advertisement of fake reviews’. This would strengthen the approach by ensuring that such sites take responsibility and work to remove the sale of fake reviews on their sites (both proactively and in response to notifications), and will in turn reduce options for review sellers to reach their customers.

Avoiding a one-size-fits-all approach for requirements on review platforms

11. The Government’s proposed additions to the CPRs has the potential to make a positive impact in addressing fake reviews, but the information provided so far is quite topline. The success of the approach will, in part, depend on the guidance the Government provides to accompany this. In particular, this will include how the guidance sets the scope with regard to product and service (retailer) reviews, if both are indeed within scope, and how it defines terms such as ‘reasonable and proportionate’ with respect to checking if reviews are genuine.

12. One of the goals of the Digital Markets, Competition and Consumers Bill is to enhance digital competition and it is vital that it does likewise in the space of online reviews by both recognising and supporting the variety of models in the landscape. Some aspects to consider in this respect include:

i) Clarifying whether reviews about products and retailers are both within scope, and if so, ensuring the approach considers their unique differences

13. In requiring platforms which host reviews to take ‘reasonable and proportionate steps’ to ensure reviews are genuine it is vital that a careful and holistic approach is taken, otherwise the review sector - and all the benefits it brings - could be undermined.

14. Often in discourse it is easy to oversimplify the review landscape based on our personal experiences. For example, when thinking about online reviews, people often envisage a retail site which sells products and displays reviews for those products on the page of the item which is for sale. Whilst this is one type of review, the review landscape is actually much more varied and vibrant with reviews.

15. The first differentiator of reviews is the type of review. These fall into two categories: (a) a product review, where consumers are reviewing a physical item, and (b) a service review, where consumers are reviewing a merchant, retailer or business who sells products or services and have as their main object the performance of the seller and their customer service level. At Trustpilot, we call the latter type of review "service reviews", but we note that the research paper refers to them as reviews of retailers. While it may seem like a technicality to distinguish between types of reviews, it is important to ensure that this difference is recognised when setting the parameters for whether a platform has taken ‘reasonable and proportionate steps’ to ensure a review is genuine.

16. For a product review, it could be sensible to require a platform to check for evidence of the purchase, such as a receipt. Yet for a service review, this would not be suitable as a service experience does not always include a purchase. For example, a customer may have such a bad experience in trying to make a purchase, such as unexpected fees appearing in the purchase experience or being passed from pillar to post on a phone call, that they abandon their purchase. In such an event the consumer has a genuine experience which is valuable for both other consumers and also for the business to hear about, but they do not have a receipt for said experience.

17. Likewise, requiring that a reviewer has used the product they write a review about makes sense for product reviews, however, in the context of service reviews it becomes more difficult to define. For example, when does a consumer ‘use’ a service? Any gaps in the definition could be exploited by businesses in an attempt to remove genuine reviews and silence consumer voices by preventing them from sharing their experiences.

18. We would note the Government’s research into fake reviews which accompanied the publication of the Bill only focuses on product reviews - "We only examined the prevalence and impact of fake review text on consumer products. As such, we cannot determine whether these findings also extend to services purchased online or other misleading review practices." [7] Despite this, there is still a conflation of the review landscape as, despite recognising that the research only covers product reviews, the paper also states: "we can be confident that our study robustly estimates the impact of fake reviews on UK consumers." [8]

19. Care should be taken to accommodate and understand different models and avoid extrapolating product review conclusions to other models without further research. As it stands, there is significant variation found in the evidence base into fake reviews:

The Government’s research (cited above) focused only on product reviews on e-commerce platforms, i.e. sites which sell products and have a review function alongside this, rather than specialised review platforms. Within this aspect of reviews, the focus was further narrowed to look at three categories of items (consumer electronics, home and kitchen, sports and outdoors). For this specific category of reviews, they estimated that 11% to 15% of the reviews were fake [9] .

The World Economic Forum meanwhile has analysed official figures and self reported figures from ‘the world’s leading e-commerce sites’ (a combination of product and service e-commerce and review platforms) and has found that on average 4% of online reviews are fake [10] .

Added to this, Trustpilot’s last Transparency Report disclosed that we identified 5.4% of reviews submitted to our platform as being fake [11] . The majority of these are identified and removed prior to being published on our site.

20. In considering the proposed text, it will be important to clearly define what steps are deemed to be ‘reasonable and proportionate steps’ for platforms to take to ensure a review is genuine. Specifically, ensuring that this definition can flex to both product and service reviews will be integral to the success of the text.

21. The issue of clarity also arose in the EU’s "Omnibus Directive" [12] which came into force in May 2022 and also sought to address fake reviews online via amendments to the Unfair Commercial Practices Directive. This Directive similarly prohibits traders from stating that reviews are submitted by consumers who actually bought or used a product, without first taking "reasonable and proportionate steps" to ensure that this is true. However, while the Directive referred to "consumer reviews of products" which clearly encompasses product reviews, the updated Guidance on the Directive [13] extended the requirements to also cover consumer reviews of traders or businesses (service reviews). In practice, we experienced that inconsistency in the wording between the Directive text and the Guidance resulted in confusion (amongst groups including the media, businesses, traders and some consumers) regarding which types of reviews were covered by the legislation, and fuelled confusion about which steps were necessary to ensure that reviews are genuine.

22. In light of this experience, we suggest that where both types of reviews are intended to be within scope, this should be made clear at the outset. Furthermore, clarity should also be provided when setting out the requirements for what are considered ‘reasonable and proportionate steps’, ensuring that the distinctions between the requirements for product reviews and service reviews are clearly conveyed.

ii) Recognising open and closed platforms

23. Not only do the types of experience which review platforms include vary, so does their wider model with regard to who can leave reviews, and how they can leave reviews. This can be viewed through the lens of ‘open’ and ‘closed’ platforms.

24. Closed review platforms require reviewers to be invited by a business to leave a review about them. Whilst this has some benefits such as the business directly drawing from its customers, it also has downsides in that the business can choose who it hears from and can filter out customers by not inviting those it knows were dissatisfied or didn’t have a good experience. Likewise, a consumer may wish to share their experience but has no opportunity to do so as the business has control of who leaves reviews and so can pre-screen and review gate who is able to submit reviews, including by controlling the invitations issued to leave reviews. Some models also allow businesses to pre-moderate reviews before they’re published. As such, the full picture is not presented for consumers reading the reviews.

25. Open platforms enable consumers to write reviews at any time, without an invitation, which broadens the types of experiences included, extending feedback from purchases only to encompass customers who decided not to buy. Trustpilot is an open platform and we recognise that openness comes with a responsibility - but platforms can adopt appropriate safeguards and Trustpilot has demonstrated what a proactive and responsible approach looks like. We have adopted a range of safeguards, such as automated fraud detection software, identity verification, investigation of flagged reviews, automated invitation services, actions against misuse and review sellers, and transparency on processes and practices - including how reviews are collected.

26. In order to maintain a competitive environment, it is important that both open and closed models are catered for in this legislative action so as to enable consumers and businesses to choose which to interact with and draw from, based on their unique needs and wishes.

iii) Enabling innovation to flourish

27. Against this backdrop of a very varied review landscape, it is important to appreciate that there is also no one-size-fits-all approach to addressing fake reviews. Whilst it is critical that platforms are brought up to a baseline standard in tackling fake reviews - and this Bill can play a key role in that - it must be done in a way that includes flexibility for platforms to apply the right types of safeguards to reflect their particular model, and encourages rather than quashes innovation.

28. As the Government itself notes, [14] fake reviews are becoming more sophisticated as those seeking to propel them work to ensure that there is little to distinguish them - visually - from genuine reviews. As we know from our own work in tackling fake reviews, they are often not discernible to the naked eye and therefore to the public. Instead, at Trustpilot we utilise a host of methods to identify fake reviews. This includes deploying carefully trained, bespoke fake review detection software that works across our platform to detect misuse and is supported by a team of dedicated fraud and investigation specialists. Every review is run through this automated technology, which identifies unusual patterns, based on multiple data points, such as IP addresses, user identifiers, device characteristics, location data and timestamps. On average our automated software analyses over 128,000 reviews every day. As we collect more and more information about fake reviews and the behavioural patterns linked to them, these systems improve in their capability to identify and tackle fakes, and the individuals behind them.

29. We are constantly innovating in this space and have also provided consumers with optional tools they can deploy, such as enabling both consumers and businesses to verify their identity. The process requires both parties to safely and securely share a copy of their government-issued photo ID, as well as take a selfie. This is supplied to and verified by a third party, using the same technology used by banks, healthcare providers and educational institutions. This is optional and allows everyone to play their part in building an even more trusted community on Trustpilot, and we are now able to highlight where businesses and consumers have done this, as a ‘trust signal’ to those looking at the information.

30. By engaging technology, people and our community, we work to tackle fake reviews and provide an annual Transparency Report to share our latest data and innovations with the public and policymakers. However, while this is the approach Trustpilot has taken to reflect that ours is an open platform which hosts service reviews, other platforms will have developed other systems and approaches which are unique to their platforms. Bad actors seeking to deploy fake reviews are constantly developing their methods and, in turn, this needs to be matched by ongoing innovation by review platforms tailored to their circumstances.

31. This legislation and the subsequent use of the powers which it grants must enable and support innovation by avoiding an inflexible approach to what it deems ‘reasonable and proportionate steps’ which can be taken by review platforms. This area is constantly evolving and, whilst it’s important to ensure that platforms are working to meet that, limiting the tools available for how that can be done via prescriptive guidance would be detrimental, both now and in the future (as technology evolves), to the end goal of improving the authenticity of online reviews.

32. It is vital that this legislation balances requirements with ensuring a healthy ecosystem and consumer choice by allowing a range of models to flourish. This facilitates the sharing of full insights that maximise the ability for consumers and businesses to help one another. Regulation should ensure that platforms put relevant safeguards in place, but must not effectively prohibit open platforms in favour of closed, nor prioritise product reviews above service reviews.

Delivering a pro-competition approach to reviews

33. As noted above, consumer reviews play an important role in driving competition as they inform consumer behaviour and also help businesses to refine their offering. Added to this, as also noted above, the online reviews world has a range of businesses and models, and this enables competition within the sector.

34. It is important that this competition is not reduced due to the approach of the Bill and the subsequent use of the powers it grants. This Bill is intended to champion digital competition, and that should also extend to how it approaches the world of online reviews. Avoiding a one-size-fits-all approach is key for the sector to flourish and continue serving consumers and businesses, whilst ever improving and competing to deliver good offerings to these audiences.

35. Against this backdrop, it is also important to protect the ability to innovate and encourage new entrants into the market. We would therefore urge that prescriptive measures are avoided. Such measures risk the ability of review platforms to innovate in the future, and also have the potential to erect barriers to new entrants to the market, as prescriptive restrictions may prove costly.


36. The Digital Markets, Competition and Consumers Bill brings significant opportunities to tackle fake reviews if the details, which are yet to be finalised, achieve the right balance. Based on the Government’s proposed approach, Trustpilot supports the focus on review sellers and sees potential in ensuring that review sites are all working to crack down on fake reviews. We would also welcome further action from the Government to extend the onus to take action to ISP and social media sites which host the sale of fake reviews.

37. It is critical that the final approach to addressing fake reviews supports and enhances competition in the online reviews sector. This can be achieved by understanding and recognising the variety in the sector, be that of review type or by business model, and avoiding a one-size-fits-all approach to setting the requirements on platforms. Competition and innovation must not be stifled.

June 2023

About Trustpilot

Trustpilot is a leading online consumer reviews platform that brings businesses and consumers together to foster trust and inspire collaboration. Our vision is to be a universal symbol of trust.

Trustpilot is free to use, open to everybody and built on transparency. We host reviews to help consumers shop with confidence and deliver rich insights to help businesses improve the experience they offer. The more consumers use our platform and share their own opinions, the richer the insights we offer businesses, and the more opportunities they have to earn the trust of consumers, from all around the world.

Founded in 2007 in Denmark, we listed on the London Stock Exchange in Spring 2021. We are a global company and the UK is our largest market. Over 190 million reviews have been submitted to Trustpilot and over 4 million reviews are posted per month, with our most recent figures showing that UK consumers submit the most reviews on our site. We have a very strong UK presence with offices in London and Edinburgh.

[1] Digital Markets, Competition and Consumers Bill - Annex 4: Impact assessment of wider measures , Department for Business & Trade and Department for Science, Innovation & Technology, April 2023

[2] Ibid.

[3] UK consumer review of Trustpilot, posted 6 June 2023 . See also: "[...] Buying online can be unsafe so reading others' reviews gives me peace of mind [...]", 8 June 2023 ; "Great to be able to leave feed back to help other people and also what others have to say about products and companies before to use them", 7 June 2023; and "[...] Great to be able to praise good , well run, honest companies as well as to be able to warn people of unworthy and unsafe companies to avoid at all costs. [...]", 24 January 2023 .

[4] UK consumer review of Trustpilot, 7 June 2023 ..

[5] UK business review of Trustpilot, 24 March 2023 : "[...] Can highly recommend to any business owner wanting to create a real connection with their customers/ audience [...]". See also: "[...] it’s great to see what customers are thinking of your products", posted 7 June 2023 .

[6] Fake Online Reviews Research: Estimating the prevalence and impact of fake online reviews , Department for Business & Trade, April 2023.

[7] Ibid.

[8] Ibid.

[9] Ibid.

[10] Fake online reviews cost $152 billion a year. Here's how e-commerce sites can stop them , World Economic Forum, August 2021.

[11] Transparency Report 2021 , Trustpilot, published 2022

[12] The Better Enforcement and Modernisation Directive (Directive (EU) 2019/2161), known as the Omnibus Directive.

[13] Guidance on the Directive interpretation and application of Directive 2005/29/EC concerning unfair business-to-consumer commercial practices in the internal market, at 4.2.4.

[14] Fake Online Reviews Research: Estimating the prevalence and impact of fake online reviews , Department for Business & Trade, April 2023, at p11, and footnote 26 on p20: "[...] fake reviews have become more sophisticated over time (as more obvious features of reviews such [as] sentiment and text are no longer good predictors of whether reviews are genuine or fake)."


Prepared 14th June 2023