14 Types of Product Reviews: Source, Format, and Platform

Product reviews are ratings, written statements, and customer-created media items that evaluate a specific product from an asserted purchase or usage experience. Product reviews fall into 14 types. The 14 types sit on three classification dimensions rather than in 14 separate boxes. The three dimensions are the source that produced the review, the format the reviewer submitted, and the platform that displays the finished review. The source dimension holds seven types. The seven are verified purchase, unverified, incentivized, sampled, syndicated, expert, and fake product reviews. The format dimension holds four types. The four are star-rating-only, written, photo, and video product reviews. The platform dimension holds three types. The three are on-site, marketplace, and Google Shopping product reviews. A single review carries one label from each dimension. A verified, written-and-photo, marketplace review is one complete classification record, and a sampled, video, on-site review is another.

A classification dimension is an axis of comparison that answers one question about a review and ignores the other two. The source axis fixes provenance and material connections. Provenance governs trust and the duty to disclose. The format axis fixes information richness. Richness governs how much a star, a sentence, an image, or a clip reveals about fit, defects, and everyday use. The platform axis fixes display rules. Display rules govern moderation, product matching, data feeds, and visibility on Amazon, on a brand's own store, or in Google Product Ratings. Amazon sellers, ecommerce teams, and direct-to-consumer brands read the three axes together for three reasons. The reasons are to build social proof, to collect user-generated content inside marketplace and Federal Trade Commission (FTC) boundaries, and to turn customer feedback into product decisions. The benefit of a clean taxonomy is a review program that counts each submission once and discloses each relationship once.

What Are the 14 Types Of Product Reviews?

The 14 types of product reviews are seven source types, four format types, and three platform types that overlap as attributes of one review. The 14 types are not 14 mutually exclusive classes. The 14 types answer three separate questions. The questions are who or what produced and supplied the review, what the reviewer submitted, and where the review is shown or delivered. The source question yields the seven provenance labels. The format question yields the four content labels. The platform question supplies the three display labels. A verified purchase label, a photo label, and a marketplace label therefore describe one Amazon review from three angles. None of the three labels contradicts the other two.

A review label is an evidence signal about origin, content, or distribution, and it is not a guarantee of truthfulness or usefulness. Source labels overlap inside their own dimension. A sampled review is also an incentivized review, since a free product has material value. A syndicated review keeps the verification or incentive status it earned on its first site. Platform labels identify the display location only. A Google Shopping rating supplied through a merchant feed is not a review authored on Google. No regulator or marketplace publishes this exact 14-item taxonomy. Amazon and Google Merchant Center apply operational labels such as Verified Purchase, star rating, photo and video, and Product Ratings. The FTC frames reviews by authenticity, incentives, relationships, and suppression instead. The table below places the 14 labels on their three dimensions beside the record each label makes and one surface where each label appears.

Why Do Product Review Types Matter For Ecommerce Brands?

Product review types matter for ecommerce brands because each type decides what a shopper infers, what the brand learns, where the content appears, and which rules apply. The source type sets the provenance signal. A Verified Purchase badge ties the review to a platform-recognized order, whereas a hidden incentive leaves the reviewer relationship open to doubt. The format type sets the diagnostic depth. A star-only rating shows direction, a written review adds reasons, a photo shows appearance and scale, and a video demonstrates motion, setup, or assembly. The platform type governs the distribution rules. On-site reviews answer to the brand's own moderation. Marketplace reviews answer to Amazon's eligibility and anti-manipulation checks. Google Shopping ratings answer to feed and product-identifier requirements.

Compliance exposure is the legal and platform liability a review program carries, and it tracks the source type. Fake reviews, rating-conditioned rewards, and concealed samples carry a different legal weight from genuine, disclosed feedback under the FTC Consumer Reviews and Testimonials Rule. Provenance also carries a measurable commercial weight. The Medill Spiegel Research Center at Northwestern University found in its 2017 study "How Online Reviews Influence Sales" that a verified-buyer badge raised purchase likelihood by 15% within its dataset. Reading the three types together lets a brand pick the content that converts, the disclosure that protects it, and the channel that reaches its buyers.

What Are the Types of Product Reviews by Source?

The types of product reviews by source are seven provenance classes that record where a review originated and which purchase, incentive, transfer, expertise, or authenticity relationship applies. The seven source types are verified purchase product reviews, unverified product reviews, incentivized product reviews, sampled product reviews, syndicated product reviews, expert product reviews, and fake product reviews. Source stands apart from format and platform. A provenance class says nothing about whether the reviewer typed a sentence or uploaded a clip. The same class is silent on whether the review sits on a brand store or an Amazon detail page. One review carries one source label at minimum and several source labels when relationships stack. Assigning a source label to a review runs through six checks, which are listed below.

1. Confirm whether the platform links the review to a recognized order, then mark the review as verified purchase or unverified without reading unverified as fake.
2. Look for money, discounts, loyalty points, free product, or a sweepstakes entry, then classify the review as incentivized and add the sampled label whenever a product was supplied.
3. Trace whether the review was collected on another site and transferred, then tag it as syndicated while preserving its original source and disclosure labels.
4. Examine whether the author is presented as a tester, publication, professional, or creator, then accept the expert label only when the expertise is relevant to the claims.
5. Test whether the asserted experience, identity, or origin is fabricated, then apply the fake label only when evidence supports that conclusion.
6. Record format and platform as separate attributes, so that one complete record reads sampled plus written and photo plus marketplace with a visible incentive disclosure.

Running the six checks in order gives every review its source label before the format and platform labels are added, and the seven source classes are listed below.

1. Types of Product Reviews by Source

Source types classify a product review by its provenance, meaning the purchase link, the material benefit, the transfer path, the author status, or the authenticity of the experience. The seven appear below in the order a classifier checks them. Each check assigns one label without canceling the labels assigned before it.

1.1 Verified Purchase Product Reviews

A verified purchase review is a source type that the platform has linked to an eligible purchase in its own transaction records. Check whether the review carries the platform's purchase badge, and classify it as verified when the badge is present. The badge proves a transaction link and nothing more. Sellers read the badge as evidence of purchase, never as proof of accuracy.

1.2 Unverified Product Reviews

An unverified review refers to a provenance class that the current platform cannot tie to a purchase in its own checkout system. Label the review as unverified when the badge is absent, and stop the purchase check there. The missing badge covers a product bought from a physical store, received as a gift, or used through work. Brands treat the label as reduced provenance rather than as an accusation of falsity.

1.3 Incentivized Product Reviews

An incentivized product review denotes a source type submitted after the reviewer received or expected a material benefit tied to leaving the review. Look for payment, discounts, loyalty points, or free products, and record any sentiment condition separately. The federal rule permits a sentiment-neutral reward for any honest review, while a reward paid only for five stars violates it even when disclosed. Marketplaces impose stricter bans than the federal rule.

1.4 Sampled Product Reviews

A sampled product review stands as an origin class written after the reviewer received a free or discounted product from a brand, retailer, sampling network, or platform program. Assign the sampled label whenever the product itself was the benefit, and keep the incentivized label too, since a product has material value. Amazon Vine is the marketplace example, and Vine Voices receive products free to give honest opinions without any promise of a rating.

1.5 Syndicated Product Reviews

A syndicated product review describes a source type collected on one approved site and displayed on another retailer, brand, or channel for the matched product. Check whether the review was transferred, and file it as syndicated with its original verification and incentive labels intact. Syndication networks match products through identifiers such as GTIN or SKU and attach a source attribution badge, because a stripped label turns a legitimate transfer into a misleading one.

1.6 Expert Product Reviews

An expert product review means a source type created by a person or organization with testing, professional, or editorial expertise relevant to the claims. Mark the review as expert only when the expertise fits the product, and log sponsorship, affiliate pay, or supplied units as separate material connections. A camera publication testing a manufacturer-supplied unit is expert and sampled at once. Expert reviewers add technical depth without guaranteeing independence.

1.7 Fake Product Reviews

A fake product review constitutes a source class that misrepresents the reviewer, the experience, the product used, or the origin of the review. Tag a review as fake only when evidence supports the finding, because a negative, anonymous, or unverified review is not fake by default. Fabricated experiences, purchased reviews, and employees posing as consumers all fall inside the class. Platforms remove the class outright, since a disclosure cannot make a fabricated experience genuine. Sorting a review into one of the seven source classes settles its provenance and leaves the submitted information unread.

2. What Are the Types of Product Reviews by Format?

Format types sort a product review by the information the reviewer submitted, from a bare score to moving images. There are four format types. The four format types follow below in rising order of information richness and collection friction. Formats combine inside one review, so a written review with two photos carries two format labels. Format never changes the source label.

2.1 Star-Rating-Only Product Reviews

A star-rating-only product review forms a format type that records a numeric score on a one-to-five-star scale without narrative text or customer media. Treat a submission as star-only when the text, photo, and video fields are empty. The low-friction format raises response volume, yet the score explains no use case, attribute, or expectation. Brands label the entry as a rating, not as testimony.

2.2 Written Product Reviews

A written product review names a content-format class containing narrative text about the product, the experience, the context, or the outcome. Classify a review as written when it carries free text, structured pros and cons, or prompt answers, with or without a star score. The text reveals reasons, fit, defects, and the language customers use. Advertisers treat reused review text as a testimonial under the federal rule.

2.3 Photo Product Reviews

A photo product review denotes a format type containing customer-submitted still images of the product, packaging, fit, result, or environment of use. Count a review as photo when at least one image is attached, and moderate the image for relevance, privacy, safety, and intellectual property. Photos show real-world scale, color, damage, and assembly that catalog images omit. Moderators keep unflattering photos as readily as flattering ones.

2.4 Video Product Reviews

A video product review ranks as the richest format class, holding customer-created moving images with audio, captions, or a demonstration attached. Record a review as video when the submission includes an unboxing, setup, test, fit check, or before-and-after sequence. Video demonstrates motion, sound, and workflow more directly than a static image, at the highest creation and moderation cost of the four. Teams accept that cost where performance must be seen or heard. Filing a review by format fixes what the submission holds and leaves the surface that displays it unassigned.

3. What Are the Types of Product Reviews by Platform?

Platform types place a product review according to the surface where it is displayed or delivered, not by the place where it was written. There are three platform types. The three platform types are set out below. Each surface runs its own solicitation, moderation, feed, and disclosure rules. A platform label identifies the display environment only, so a brand-site review becomes on-site at origin and syndicated at a retailer without changing its author.

3.1 On-Site Product Reviews

An on-site product review is a platform type shown on a brand's or retailer's own product page or controlled web property. Place a review in the on-site class when the site owner collects it directly or receives approved syndicated content and displays it. The owner controls moderation, sorting, and presentation, which carries a duty to avoid selective suppression. Retailers display source and incentive labels beside every on-site review.

3.2 Marketplace Product Reviews

A marketplace product review describes a display class published inside a third-party marketplace under that marketplace's eligibility, verification, moderation, and display rules. Classify a review as marketplace when the marketplace, not the brand, applied the labels and the checks. Marketplace reviews reach shoppers at a high-intent purchase point and supply platform-native trust signals. Sellers request reviews only through the marketplace's approved contact methods.

3.3 Google Shopping Product Reviews

A Google Shopping product review counts as a distribution type represented through Google Product Ratings in eligible Shopping ads and product listings. Assign the Google Shopping label when a merchant or approved aggregator supplied the review through a product-review feed and Google matched it to a GTIN. Google aggregates the reviews rather than authoring them, so the original verification and incentive status stays relevant. Google Merchant Center requires at least 50 reviews in the feed before ratings appear. Merchants treat the star display as distribution, not as new collection.

Stacking source, format, and platform together lets a brand record one review as three accurate labels instead of counting it as three unrelated reviews. Every label in that record attaches to a review that a customer authored.

Are AI Review Summaries Among the Types Of Product Reviews?

No, AI review summaries are not among the types of product reviews, since a summary is a derived presentation layer built from existing reviews. A customer-authored review holds one reviewer's rating, words, photos, or video, while an AI summary compresses patterns across many such reviews into generated text. Every attribute the summary shows therefore belongs to the input reviews.

Amazon's review highlights are the clearest official example. The feature draws on text-based reviews from verified purchases, identifies shared positive, neutral, and negative themes, and lets shoppers open the supporting reviews. The summary inherits the quality, balance, and recency of those reviews and adds the risk of lost nuance. AI-generated text submitted as a first-person customer experience is not a summary at all. It is a fake product review.

Which Types Of Product Reviews Carry Compliance Risk?

Compliance risk in product reviews is a legal and platform exposure created by how a review is obtained, labeled, or displayed. The format the reviewer chose carries no risk on its own. There are seven prohibited review practices and four disclosure-dependent review practices. The prohibited practices are fake reviews, sentiment-conditioned incentives, concealed incentives and samples, undisclosed insider reviews, selective suppression and intimidation, misleading syndication and attribution, and manipulated ratings or provenance. The disclosure-dependent practices are sentiment-neutral incentives, product sampling, expert and creator reviews, and syndicated reviews, which are briefly discussed below.

Fake Reviews

A fake review is a prohibited practice under the FTC Consumer Reviews and Testimonials Rule, which took effect on 21 October 2024. The rule bars reviews that misrepresent the reviewer, the experience, or where the content originated. The UK Digital Markets, Competition and Consumers Act 2024 names the same conduct a fake consumer review from 6 April 2025. Amazon enforces a zero-tolerance policy against the same abuse on its marketplace.

Sentiment-Conditioned Incentives

A sentiment-conditioned incentive is a payment, discount, or reward offered for a positive or negative rating. The FTC rule devotes a section to buying positive or negative reviews. Disclosure does not cure the condition, since a five-star requirement stays prohibited even when the reward is announced. The condition attached to the reward, rather than the reward itself, creates the violation.

Concealed Incentives and Samples

A concealed incentive is a material benefit that the published review hides from the shopper. UK law addresses publication that conceals the fact a review has been incentivized. FTC endorsement guidance requires a clear disclosure of any material connection a reasonable consumer would not expect. A free or discounted sample falls under the same connection rule, so a hidden sample carries the same exposure as hidden money.

Undisclosed Insider Reviews

An undisclosed insider review is feedback written by an owner, officer, manager, employee, agent, or close relative who poses as an independent consumer. The FTC rule and its guidance treat the insider relationship as a material connection. The relationship becomes a violation when the review omits it. A visible relationship label removes the deception without removing the review.

Selective Suppression and Intimidation

Selective suppression is the blocking of negative reviews because of their sentiment, or the display of a distorted selection of reviews. Intimidation covers threats directed at reviewers. The FTC rule and the UK statute both address suppression. European Commission guidance on the Unfair Commercial Practices Directive treats selective publication or removal as a misleading practice. Amazon adds a seller boundary, since sellers must not ask customers to remove a negative review or post a positive one.

Misleading Syndication and Attribution

Misleading syndication is the transfer of a review to the wrong product, variant, seller, or market. The practice also covers stripped source or incentive labels and duplicated reviews that inflate review volume. A destination page that implies it collected borrowed reviews falls under the same practice. Risk rises when provenance fails to travel with the content, since the shopper then reads a borrowed review as a native one.

Manipulated Ratings and Provenance

Manipulated provenance is the alteration of a review score or the fabrication of a trust signal. Examples include edited star values, star-only ratings presented as written testimonials, fabricated Verified Purchase-style badges, and a company-controlled review site presented as independent. The FTC rule addresses company-controlled review sites and fake indicators directly. European Union guidance also prohibits altering reviews and paying for fake ones.

Sentiment-Neutral Incentives

A sentiment-neutral incentive is a reward offered for any honest review regardless of its rating. The FTC rule does not prohibit an incentive merely because it is offered for a review, provided the reward carries no express or implied sentiment condition. The FTC Act and the Endorsement Guides still require disclosure of the reward. A marketplace keeps the right to ban the practice outright.

Product Sampling

Product sampling is a disclosure-dependent practice in which a free or reduced-price product is supplied in exchange for an honest evaluation. The supplied product is a material connection, so the sample relationship must appear beside the review. A compliant program requests feedback without rating conditions and follows the destination platform's approved process. Amazon Vine is the sanctioned Amazon route, not a license for off-platform sampling on Amazon.

Expert and Creator Reviews

An expert or creator review is a conditional practice because relevant expertise adds value while a hidden relationship removes it. Sponsorship, affiliate compensation, employment, samples, or special access must be clear whenever the relationship is material to the audience's evaluation. The qualification of the reviewer answers one question and the connection of the reviewer answers another. Both answers belong on the page.

Syndicated Reviews

A syndicated review is a lawful redistribution when the product match, original source, verification status, incentive label, and duplicate handling stay accurate. The practice turns risky only when those attributes are lost at the destination. Correct attribution therefore separates legitimate reach from misleading syndication. Reading these eleven practices by their condition rather than their format lets a brand separate the reviews it must refuse from the reviews it must label. That separation comes before any sentiment is read.

What Are the Types Of Product Reviews By Sentiment?

Sentiment is an analytical attribute of a product review that records the direction of the evaluation, separate from source, format, and platform. There are four types of product reviews by sentiment. The types are positive, negative or critical, neutral, and mixed or balanced reviews. A positive review expresses satisfaction, recommendation, or a favorable judgment on the product or on one attribute. A negative review expresses dissatisfaction, failure, or non-recommendation and identifies defects, expectation gaps, return drivers, and safety concerns. A neutral review describes facts, trade-offs, or an experience without a clear overall verdict, and supplies the specifications, fit, and compatibility that a star average omits. A mixed review contains favorable and unfavorable evaluations across different attributes and shows the trade-offs that shoppers with different priorities weigh.

Aspect-level sentiment is the evaluation attached to one attribute inside a single review. A buyer praises build quality, criticizes setup, and stays neutral on price within the same text. Star rating and written sentiment diverge for the same reason. Accidental clicks, changed opinions, category norms, sarcasm, or a low score for one decisive attribute pull the number away from the words. Sentiment analysis therefore preserves the overall label and the attribute-level evidence together, and that preserved detail decides which review types serve a brand best.

What Types Of Product Reviews are Best For Ecommerce Brands?

The best type of product review for an ecommerce brand is a portfolio rather than a single label. The portfolio is built from verified written reviews, category-relevant photo and video reviews, disclosed sampling, correctly attributed syndication, and channel-native coverage on the brand site and the marketplace. Each member of the portfolio answers a different business question. The verified written review combines a purchase provenance signal with reasons, use cases, and issue detail. The verified photo review fits apparel, beauty, home, food, and packaging, where scale, color, fit, and finish need to be seen. The verified video review fits assembly, tools, electronics, and sound, where process and performance need motion. Disclosed sampling serves new products and defined tester cohorts. Attributed syndication serves multi-retailer catalogs and distributed product pages.

The trade-off of the portfolio is friction and moderation burden, which rise from star-only ratings through text to video, while compliance risk rises wherever a disclosure is missing. Category fit decides the format, distribution decides the platform, and provenance decides the source. The decision rule for a brand is to maximize genuine provenance, decision-relevant detail, category fit, and compliant distribution. The highest star average or the largest review count is not the rule. A high average built on filtered criticism raises the question of whether negative reviews hurt sales at all.

Do Negative Types Of Product Reviews Hurt Sales?

Yes, negative product reviews hurt sales, but only when the criticism signals a functional, service, or safety risk that a shopper cannot discount. A repeated defect complaint lowers purchase probability and redirects demand to substitutes, while a taste complaint reveals fit for one segment. Varga and Albuquerque found in "The Impact of Negative Reviews on Consumer Search and Purchase Decisions" in the Journal of Marketing Research in 2024 that the effect is stronger for utilitarian information than for hedonic claims. Suppressing the criticism breaches FTC, UK, EU, and Amazon rules, so the lawful remedy is a response through permitted channels, and that response feeds the trust a review earns.

Do All Types of Product Reviews Carry Equal Trust?

No, product reviews do not carry equal trust, since trust depends on ten factors rather than on one badge or one format. The factors are purchase verification, material-connection disclosure, genuine experience, relevant expertise, specificity, recency, balance, source attribution, platform safeguards, and consistency with other evidence. A Verified Purchase label is a stronger provenance signal than no transaction link, yet it certifies a transaction and not the accuracy of the opinion. A disclosed sampled review is more transparent than an organic-looking review with a hidden relationship. A mixed rating history reads as more trustworthy than uniformly extreme scores.