Online reviews

Are consumers losing faith in product reviews

As seen on The Future of Commerce Blog:

There are many occasions where we will all rely on reviews to help us distinguish between products or services; especially when they have many different choices.

I recently bought a digital thermometer on Amazon where I had the choice of many hundreds of devices so I relied on the reviews to help me make the decision.  Each time I used it, I would get completely random results and it was clear that the device was not as reliable as the reviews had led me to believe.  A few days later I received the normal email from Amazon asking me to review it so I posted an accurate and honest negative review.  I often review my Amazon purchases as I know that I tend to rely on them myself when buying from the website.  Like many customers, I tend to review the items I really like and those that I really don’t.

What happened next was quite an eye opener into the world of Amazon reviews.  Over the following few weeks, I received many emails from different addresses, all offering me money to remove the negative review.  They even provided a link with instructions on how to do it.  Despite me replying stating that I was unwilling to do this, the emails continued.  It only stopped once I reported it to Amazon, who, to their credit, took it quite seriously.

This got me thinking. If this has happened to me, how many other people are tempted by the offer of money and go ahead and remove a negative review?  How about the positive reviews; do Amazon sellers pay people to post them?  Can we even really trust reviews at all?

Countless studies have shown that user-generated reviews can have a big impact on your conversion rates, although, admittedly, many of these studies have been published by companies that sell review management services.  It completely stands to reason that user-generated reviews can be very powerful tools in influencing a buying decision; especially with something technical or expensive such as a new TV or a holiday.  I would expect that almost everyone reading this post has viewed Amazon, Yelp or TripAdvisor reviews at some point and these have influenced their buying choices.  With many marketplaces that are flooded with different product options, many at a similar price, reviews can be a key differentiator.  But is the race to win on reviews damaging the trust customers have in them and will this, ultimately, dilute their importance and the impact they have on conversion rates?

Man sitting with his laptop using a mobile phone

Bulk-buying reviews

A recent investigation and report by Which? detailed how easy it is for Amazon sellers to buy verified reviews in bulk.  Just a simple Google Search finds a number of companies offering a review generation service.  They use techniques to get around the Amazon guidelines and to ensure that the reviews appear to be genuine.  Most of the time the process involves providing your product to reviewers for free or at a discounted rate but, for an additional fee, some services will guarantee that your product is returned once the review is submitted.  As well as product reviews on marketplaces such as Amazon or Walmart, sellers can also purchase seller feedback and even votes to push positive reviews up.  The prices for these services start from around $10 per review but can be much higher for certain categories of products which highlights how valuable genuine-looking reviews can be.

During their study, the Which? researchers posed as an Amazon seller to speak to AMZTigers, a company based in Germany, which claims to have over 60,000 reviewers worldwide to gain an understanding of what their service offers.  As well as offering to generate product reviews, the AMZTigers account manager claimed that they can even help a seller gain ‘Amazon Choice’ status within 2 weeks.  This is a big consideration as the Amazon Choice label is another big differentiator and it can possibly give the impression that the product is endorsed by Amazon.  While Amazon won’t say exactly how an item is awarded ‘Amazon Choice’ status, it has said that the item has to be highly rated, well-priced and available to ship immediately.

The fact that sellers are willing to pay large sums of money and give away free products in exchange for positive reviews is a good indication of how powerful positive reviews are.  If not, companies like AMZTigers would not exist.

Amazon’s review policies

Amazon has strict policies aimed at protecting the integrity of reviews.  They claim to analyse over 10m reviews every week and to regularly ban, suspend or even take legal action against those that violate those policies.  In my own case, as soon as I reported the seller to Amazon, the emails offering to pay me to remove the negative review immediately stopped so I am assuming that Amazon did something about it – although the product is still for sale on Amazon from the same seller.

But is Amazon doing enough?  Are they protecting the buyers they rely upon?  Companies offering these services can be easily found on search engines so it is hardly a clandestine underground service. I suspect many of us have previously read reviews for an Amazon product that are clearly for a different product or just don’t appear real to know that Amazon’s policies are not preventing this issue from happening.  While the problem is not entirely limited to Amazon, it is one of the biggest marketplaces in the world so deserves a lot of scrutiny.

What you can do to spot the fakes

So how can users trust reviews they see on websites like Amazon, eBay or Walmart?  We know that reviews can have a high impact on a user’s purchase decision so when a merchant has gamed the system and generated a lot of suspect reviews, the customer is the one that loses out.

The key to weeding out suspect reviews is learning how to recognise reviews that are likely to be fake.  Red flags include products that have a lot of very recent positive reviews, a large number of reviews that have photos attached to the review (how many people really do that?), overly detailed reviews for products that really don’t deserve such a write up, reviews that state that the reviewer was given the product free of charge and reviews that are clearly for a different but related product; known as review merging.

Another option is to use a tool such as Fakespot or ReviewMeta.  These are both very useful tools that use multiple algorithms to analyse reviews for products to provide the user with an overall score for their accuracy and reliability as well as providing more detailed insight into the patterns that they have discovered.  They can even provide an adjusted review rating, which often remove approximately 1 star from the rating.    The tools look for patterns which indicate suspect reviews without the user having to trawl through thousands of reviews.  Some of these tools even provide a browser plugin that overlay the review score on Amazon product listing and display pages which can be extremely useful.  While these tools will not be foolproof, the score and insight they provide should be taken into consideration when making a purchase decision.

The very fact that these tools exist, illustrates that there is a widespread issue with sellers gaming reviews

 

The future of reviews

So will mistrust in reviews start to erode their value and will customers stop using them as part of their purchase decisions?  Without action, I think the answer is: Yes.  While this is unlikely to be a sudden change, the slow erosion in trust will continue to lessen the impact that reviews have on conversion rates and their value. Trust is such an important aspect in customer experience.  It takes a long time to gain and a short time to lose. Perversely, as the value of gaming the system reduces, sellers will be less likely to pay money to obtain fake positive reviews and therefore they may become more trustworthy again but this would be hard to predict.

Ultimately, the solution is for the marketplace platforms to do much more to prevent this happening.  I do not believe that my experience was unique and the fact that companies like AMZTigers, Fakespot and ReviewMeta even exist and openly advertise their services tells us that a problem exists.  As trust in marketplace reviews diminishes, this will have a knock-on impact on consumer trust in the marketplace itself.  If I can’t trust reviews on Amazon, I am less likely to buy from them so resolving this issue is not just the right thing to do, but it could have a long-term commercial impact.

Marketplaces should dedicate more resources to preventing the gaming of their review programmes as well as making the consequences for breaching the terms more severe.  In my case, the seller clearly and brazenly breached the terms but is still selling the same product on the platform. Only where the cost and consequences of gaming the system outweigh the benefits will sellers stop.

More Insights

You may also be interested in...

UX is part of UI is part of CX

Customer Experience

Create customer experiences that deliver brand loyalty

At KPS, we are on a mission to simplify the buying journey and make your customers' experience more enjoyable. Our unified commerce approach delivers a consistent customer experience across all touchpoints, providing you with an intelligent, integrated and scalable platform for managing all customer interactions. 
Friction v frictionless checkout: someone using a laptop

Read more

Friction vs Frictionless: What is the right amonth of friction in commerce?

On paper, friction in any online journey is generally considered a negative thing. UX consultants strive to remove friction from online checkouts to ensure that customers work as little as possible in order to purchase a product.  Convention tells us to remove as much friction as we possibly can. But is there such a thing as having too little friction? 
Brakes webshop on a laptop

Case Study

Brakes: A new B2C site, from idea to launch in just one week

Within two days, supported by e-commerce specialists KPS, Brakes opened its B2B site to the public, accepting phone orders for collection from a select few distribution centres (DCs). Brakes recognised that in order to support its communities, it would also have to launch a complementary Direct to Consumer (D2C) e-commerce site, which would support the click and collect option. KPS was the obvious partner of choice.

Privacy settings

We use cookies on our website. Some of them are essential, while others help us to improve this website and your experience. With your consent, we use cookies to analyse the usage of our website.

We also use cookies for marketing to help us measure the success of our marketing efforts. In the settings you will find detailed information about the individual cookies, and you can refuse the use of cookies.

You can change or revoke your selection at any time on any KPS.com page in the footer under Privacy Settings.

Essential Cookies

Essential cookies enable basic functions and are necessary for the proper functioning of the website.

Cookie Banner

  • Cookie Name: kps-privacy-settings
    Value: PrivacyLevel_0#a
    Storage Duration: 4 weeks
    Description: This cookie is required to save cookie settings. It is a purely technical cookie.

HubSpot:

  • Cookie Name: __hs_opt_out
    Storage Duration: 6 months
    Description: This cookie is used by the opt-in privacy policy to prompt the visitor to accept cookies again. It is set when you allow visitors to disable cookies. It contains the string "Yes" or "No."

  • Cookie Name: __hs_d_not_tracking
    Storage Duration: 6 months
    Description: This cookie can be set to prevent the tracking code from sending information to HubSpot. It contains the string "Yes."

  • Cookie Name: _hs_initial_opt
    Storage Duration: 7 days
    Description: This cookie is used to prevent banners from being displayed every time visitors access your website in strict mode. It contains the string "Yes" or "No."

  • Cookie Name: __hs_cookie_cat_pref
    Storage Duration: 6 months
    Description: This cookie is used to capture the categories a visitor has consented to. It contains data about those categories.

  • Cookie Name: __hs_gpc_banner_dismiss
    Storage Duration: 3 months
    Description: This cookie is used when the Global Privacy Control banner is dismissed.

  • Cookie Name: hs_ab_test
    Storage Duration: Until the end of the session
    Description: This cookie is used to ensure that visitors see the same version of an A/B test page they were shown previously. It contains the ID of the A/B test page and the ID of the variant selected for the visitor.

  • Cookie Name: __cfruid
    Storage Duration: Until the end of the session
    Description: This cookie is set by HubSpot's CDN provider in accordance with its rate-limiting policies.

  • Cookie Name: ___cfuvid
    Storage Duration: Until the end of the session
    Description: This cookie is set by HubSpot's CDN provider in accordance with its rate-limiting policies.

  • Cookie Name: ___cf_bm
    Storage Duration: Until the end of the session
    Description: This cookie is set by HubSpot's CDN provider in accordance with its rate-limiting policies.

Functional cookies

Functional cookies are used by third parties or publishers to display personalised advertisements. They do this by tracking visitors across websites.

Google Analytics

Cookie Name: _ga
Storage Duration: 2 years
Description: Used to distinguish between users.

Cookie Name: _gid
Storage Duration: 24 months
Description: Used to distinguish between users.

Cookie Name: _gat
Storage Duration: 1 minute
Description: Used to throttle the request rate. If Google Analytics is provided via Google Tag Manager, this cookie is named dc_gtm<property-ID>.

Cookie Name: hide-cookieinfo
Storage Duration: 30 days
Description: Cookie set for querying the cookies.

HubSpot

Cookie Name: __hstc
Storage Duration: 6 months
Description: The main cookie for visitor tracking. It contains the domain, the user token (hubspotutk), the first timestamp (of the first visit), the last timestamp (of the last visit), the current timestamp (for this visit), and the session count (increases with each subsequent session).

Cookie Name: hubspotutk
Storage Duration: 6 months
Description: This cookie tracks the identity of a visitor. This cookie is passed to the HubSpot software when submitting a form and is used for deduplicating contacts. It contains an opaque GUID to represent the current visitor.

Cookie Name: __hssc
Storage Duration: 30 minutes
Description: This cookie tracks sessions. It is used to determine whether the HubSpot software needs to increment the session count and timestamps in the __hstc cookie. It contains the domain, the number of page views (viewCount, increases with each page view in a session), and the session start timestamp.

Cookie Name: __hssrc
Storage Duration: End of session
Description: Whenever the HubSpot software changes the session cookie, this cookie is also set. It determines whether the visitor has restarted the browser. If this cookie is not present when HubSpot manages cookies, it is considered a new session. It contains the value "1" if present.

External media

Content from video platforms and social media platforms is blocked by default. If cookies from external media are accepted, access to this content no longer requires manual consent.

LinkedIn Tag

Cookie Name: bcookie
Storage Duration: 365 days
Description: Browser Identifier cookie to uniquely identify devices accessing LinkedIn to detect abuse on the platform and for diagnostic purposes.

Cookie Name: bscookie
Storage Duration: 365 days
Description: Used for remembering that a logged-in user is verified by two-factor authentication and has previously logged in.

Cookie Name: JSESSIONID
Storage Duration: Session
Description: Used for Cross-Site Request Forgery (CSRF) protection and URL signature validation.

Cookie Name: lang
Storage Duration: Session
Description: Used to remember a user's language setting to ensure LinkedIn.com displays in the language selected by the user in their settings.

Cookie Name: lidc
Storage Duration: 1 day
Description: To facilitate data center selection.

Cookie Name: sdsc
Storage Duration: Session
Description: Signed data service context cookie used for database routing to ensure consistency across all databases when a change is made. Used to ensure that user-inputted content is immediately available to the submitting user upon submission.

Cookie Name: li_gc
Storage Duration: 180 days
Description: Used to store consent of guests regarding the use of cookies for non-essential purposes.

Cookie Name: li_mc
Storage Duration: 180 days
Description: Used as a temporary cache to avoid database lookups for a member's consent for use of non-essential cookies and used for having consent information on the client side to enforce consent on the client side.

Cookie Name: UID
Storage Duration: 720 days
Description: Cookie used for market and user research.

Cookie Name: UserMatchHistory
Storage Duration: 30 days
Description: LinkedIn Ads ID syncing.

Cookie Name: AnalyticsSyncHistory
Storage Duration: 30 days
Description: Used to store information about the time a sync took place with the lms_analytics cookie.

Cookie Name: lms_ads
Storage Duration: 30 days
Description: Used to identify LinkedIn Members off LinkedIn for advertising.

Cookie Name: lms_analytics
Storage Duration: 30 days
Description: Used to identify LinkedIn Members off LinkedIn for analytics.

Cookie Name: li_fat_id
Storage Duration: 30 days
Description: Member indirect identifier for Members for conversion tracking, retargeting, and analytics.

Cookie Name: li_sugr
Storage Duration: 90 days
Description: Used to make a probabilistic match of a user's identity.

Cookie Name: _guid
Storage Duration: 90 days
Description: Used to identify a LinkedIn Member for advertising through Google Ads.

Cookie Name: BizographicsOptOut
Storage Duration: 10 years
Description: Used to determine opt-out status for non-members.

Cookie Name: li_giant
Storage Duration: 7 days
Description: Indirect identifier for groups of LinkedIn Members used for conversion tracking.

Cookie Name: oribi_cookie_test
Storage Duration: Session
Description: To determine if tracking can be enabled on the current domain.

Cookie Name: oribili_user_guid
Storage Duration: 365 days
Description: Used to count unique visitors to a website.

Cookie Name: ln_or
Storage Duration: 1 day
Description: Used to determine if Oribi analytics can be carried out on a specific domain.

Cookie Name: ar_debug
Storage Duration: Session
Description: Supports Google attribution reporting API integration to mitigate signal loss.

Lead Forensics

Cookie Name: __lfluuid
Storage Duration: 10 years
Description: The _lfuuid cookie allows a website to track visitor behavior on the sites on which the cookie is installed. Tracking is performed anonymously until a user identifies himself by submitting a form.

Changing privacy settings

Imprint

Data protection statement