By most any measure, product reviews are one of the most important types of content. Audiences spend serious amounts of time consulting reviews of products, relying on them to make decisions. Some devote even more time to reviews, expressing their own views about products and services.
Despite their obvious importance, there is little consensus within content marketing precisely how product reviews matter to audiences. Content marketers tend to focus inwardly on their own interests, rather than the audience’s. They want to talk about their own products, and ignore the existence of competitors. Depending on their skills and inclinations, content marketers will emphasize the role of owned media (if favoring branded content), social media (if favoring social media influence and reputation management) or promoted media (if relying on PR or promotional favors such as special access). Yet such attempts at influencing customer opinion of products are of little value unless one first truly understands audience needs.
Brands need to embrace an audience centric perspective toward product reviews if they want to sell more online and become a trusted source of product information. Audiences don’t just consult product reviews to evaluate products; they actively evaluate the information in those reviews to determine if relevant to their situation. The discipline of content strategy can help brands identify the different elements audiences seek from reviews, and what information brands need to deliver.
The significance of reviews
The importance of product reviews continues to grow. For nearly every product category, people are buying more online each year. For some categories such as travel, more than half of purchases are made online in some countries. Even with more localized products such as groceries, online purchasing is increasing. Buying online requires confidence that one is buying the right product or service. Consumers are also spending more on digital products and services, which also require review information. Even when people buy a product or service locally, they rely on online reviews to make decisions. Time is limited, while choices proliferate.
The vast quantities of product review content need to be managed appropriately. This content can be enhanced to make it more useful to audiences to support their buying needs. However, brands for the most part have neglected to do this. Most rely on a simple template of allowing their customers to rate their products with stars, and leave some comments in an open text field. The benign neglect of product review content results in an unsatisfying customer experience — the information is not as helpful as it could be. It hurts the brand hosting the review, because they provide information that doesn’t really answer the needs of audiences.
Product review content is a distinct genre with a long established history. In many respects, online product reviews are less helpful than their pre-digital ancestors. To understand the potential of the product review genre, I will draw on an extensive study of reviews by Grant Blank: Critics, Ratings, and Society: The Sociology of Reviews. Blank’s book, which was published in 2007, barely discusses online reviews, but instead provides a very detailed examination of reviews by newspapers, magazines, and various kinds of consumer surveys. His insights provide rich material for rethinking how online reviews should be managed.
Basic types of reviews
Blank categorizes reviews into two basic types: connoisseurial, and procedural. They are different approaches, and each has unique strengths.
Connoisseurial reviews reflect the skills and knowledge of an individual reviewer. For a connoisseurial review to have impact, the reviewer must have credibility with the audience. Readers assume the reviewer knows what she is talking about because she has written other reviews on the same type of product or service, and has built a reputation as someone with deep knowledge who can be relied on. A connoisseur’s impact is measured by how much they influence their audience, or possibly how much they influence the producer of the product. Sometimes a reviewer’s impact is so great that they attract a regular audience of readers who may not even be looking to purchase a product, but enjoy hearing what the reviewer has to say, because they share interesting insights that add to the reader’s understanding of a topic. Connoisseurs don’t try to review everything, only the stuff that’s most notable.
The classic kinds of connoisseurial reviewers are the great restaurant, theater and movie critics. People like Craig Claiborne reviewing New York restaurants, or Roger Ebert reviewing movies. Their reviews reached many people, and could have a big impact on the suppliers of the products they reviewed. Some connoisseurial reviews are done by a corporate entity known for their expertise in an area, such as restaurant reviews by Michelin (France) or Gambero Rosso (Italy).
In the post-mass-market, digital age, it is tempting to believe that connoisseurial reviews no longer play a big role, but this would be a mistake. Some connoisseurial reviewers started in legacy media but have moved entirely to the digital world. People like David Pogue and Walt Mossberg reviewed technology gadgets for influential newspapers before starting their own blogs. On a smaller scale, connoisseurial reviewers are evident in many places. In this age of self-branding, people want to show off their knowledge. Members of LinkedIn and Quora answer questions posted, often in the hope of building their reputations. Numerous platforms cater to the output of bloggers and video bloggers who comment on the offerings of fashion, technology, and products for children. Some of these bloggers and video bloggers have developed enormous followings that rival the reach enjoyed by legacy-media critics.
Procedural reviews reflect the results of tests on a product. They generally compare several similar products, and note the differences between them. The tests are meant to be transparent, and reliable, based on uniform criteria — the same test will yield the same result no matter who conducts the test. There is an emphasis on developing data on a number of attributes of a product, and converting these data points into a numerical score that audiences will consider objective. Because they compare many products with complex attributes, sometimes they yield surprising results. The purpose of compiling all this detail is to support the purchase decisions of readers.
The archetypal examples of procedural reviews are the product evaluations of Consumer Reports (US), or Which? (UK.) PC magazines did extensive procedural reviews of computer related products, with multicolumn data tables comparing features and performance results. Among digitally native publishers, procedural reviews are less common, but some specialist sites will review products to test them for their real battery life, or their shock resistance.
While Blank considers connoisseurial and procedural reviews the two main categories of reviews, he acknowledges some hybrids that often involve surveys. A notable example of a hybrid is a Zagat guide, which combines a procedure for reviewing restaurants with the judgments of different individuals who act as dining critics. Zagat was the first product review to utilize public opinion surveys of customers. Another prominent survey of critics is the Academy Awards.
Other types of hybrid review include those that that use a procedural process to evaluate products based on their historical performance. Morningstar does this for financial products, while Consumer Reports surveys car owners to get warranty and repair data. The popular but controversial university rankings by US News also combine survey and performance data. Yet another kind of review-like listing involving surveys are lists that rank products according to their popularity. Popularity reviews include the Billboard charts, the New York Times bestseller list, and box office charts for films. The presumption of such lists is that what is most popular is what is best.
What makes a Quality Review?
Reviews that are beneficial to audiences must be credible, useful, and timely, no matter what approach is used to construct the review.
Audiences consider the credibility of reviews as essential. Brands cannot presume that a review will be read at face value. Audiences can spend a lot of effort unpacking the meaning of reviews. They look for two qualities: that the reviewer is disinterested (in the sense that he doesn’t have a financial interest in the outcome of the review), and that the reviewer is knowledgeable. Both these topics are concerns with user-generated reviews. Audiences don’t know exactly who is making comments or writing reviews online. There have been numerous accusations of firms either writing fake favorable reviews of their products, culling bad reviews about them, or sabotaging rivals with bad reviews. Sometimes firms do this directly (in one widely reported case, the CEO of Whole Foods trashed a competitor anonymously) but other times firms hire surrogates to write reviews beneficial to a particular brand. Even when audiences are convinced a review is written by a “real” customer, they may still have questions as to how much that customer knows what they are talking about, and how reliable their judgment is.
Even when the review is credible — devoid of obvious bias or misunderstanding — the review is not necessarily useful. The utility of a review depends on the goals of the reader consulting the review. Blank identifies two major goals audiences have for reviews: to make decisions, and for learning and enjoyment. Some readers need very specific information to make purchase decisions; some want an overall judgment rendered, while others simply want to feel they understand what’s important about a product, and what developments are happening. The differing goals point to different qualities in reviews: one focused on granular detail, the other highlighting big themes and trends.
Finally, audiences want to feel that reviews are up-to-date, and that the information will not be rendered obsolete soon. They look for clues that something has changed: perhaps a slight model change, a different source now making the product, or indications that service quality has deteriorated or suddenly improved. Audiences are sensitive to changes, and believe that reviews should indicate a consistent experience with the quality of a product. Brands are expected to be consistent, and signs of inconsistency are worrying. The collective body of opinion reflected through different reviews is meant to help audiences predict how a product will perform for them in the future. One can see this phenomenon in the reviews of apps in an app store. A new version of a popular app is released, and suddenly reviews turn negative. Has the new version betrayed the vision of the prior versions, or is this grumbling a temporary product glitch that will be soon corrected?
While credibility, usefulness and timeliness seem like obvious standards for reviews, they can be challenging to realize under a decentralized, crowd sourced model of relying on review content that is user directed and generated.
What reviews say about products, reviewers, and readers
Reviews reflect objective and subjective qualities of products, reviewer orientation and bias, and audience preferences. It can be hard to untangle the interplay between these elements.
About the product
The first difficulty is knowing exactly what precise product the review addresses. Products and services are always changing, and sometimes these changes introduce uncertainties. Have the tech specs or has the product offer changed? Is the product now made by a different supplier, or using cheaper materials? Has a defect been fixed, or is it a random, continuing problem? Was the delivery experience good or bad because of time of year? Was a hotel stay bad because of a manager who has now been replaced? Has the priced changed, and accordingly expectations have changed as a result?
Brands sometimes make changes without changing the product name or model. Other times, they introduce new product models for minor variations, and consumers become confused if their experience with a product they bought is relevant to the model currently for sale. Not only do reviewers talk about past and current experiences, they may be inclined to speculate about future models or offers.
When comparing products, a difficulty can arise in deciding what features are considered essential to a product category. Is having a choice of color an essential quality on which to judge a product? Products with more features can appear more capable, but are not necessarily “better.” All watches tell time, and some watches do much more than that. Figuring out which watches are comparable and should be reviewed together may involve some arbitrary decisions.
About the reviewer(s)
Readers typically know little about who the reviewer is, and what motivates them. Many people have a suspicion of reviewers who seem overly enthusiastic or negative, which can reflect either a personality bias unrelated to the product (e.g., agreeability or snarkiness), or a naivety about what is reasonable to expect. Positive reviews can reflect post hoc rationalizations justifying a purchase, and negative reviews can reflect buyer’s remorse. Reviews can be as much about the reputation of the reviewer as about the reputation of the product. In France recently, in an ultimate face-off of reputations, a reviewer of a restaurant was successfully sued for damaging the reputation of that restaurant.
In addition to whether reviewers have reasonable emotional expectations, readers wonder about the reviewer’s knowledge of the product category, and whether that knowledge is appropriate for their needs. Reviewers may be expert users of a product or novice users, and may be either brand loyalists or first-time customers of the brand. Each condition carries its own set of expectations. Experts may criticize entry-level products as inferior; novice users can be wowed by mundane products, or possibly overwhelmed by them. In some product categories, people apply different frames of reference to evaluate a product. For some people the service at a restaurant is most important, for others, the authenticity of the food. Different audience segments often rate products differently, applying different criteria. When brands don’t appreciate these differences, the reviews become jumbled. Yelp reviews tend to average three stars because everyone tends to focus on different characteristics, which all manage to cancel each other out.
About the reader
Like reviewers, readers have different priorities, which can change in different situations. Sometimes an individual may prioritize convenience; other times he or she may prioritize price or features. Readers will generally have more enduring preferences about products, which will shape their preferences toward reviews. Broadly speaking, some people evaluate products (or categories of products) on a rational basis (cost/performance), and others on an emotional one (how it makes them feel). But such distinctions are less clear than might first appear. Most people choose clothing for emotional reasons, but will avoid a purchase if it doesn’t fulfill expected cost or performance criteria, perhaps learning that it looks great when new, but doesn’t hold up. Some product categories, such as hiking and biking gear, appear to be about performance criteria, but these criteria are often a means of self-expression rather than utilitarian need. When the product is truly experiential, perhaps an online course or digital music, it can be harder for individuals reviewing content to rely on explicit criteria. Instead, they are more likely to try to compare the content with something else they have experienced previously, or to rely on the judgment of others who they feel resemble them. In the magazine era, people who read certain magazines could be expected to judge products in similar ways, because they shared a common point of view about products, what was important and how to judge that.
Implications for digital content strategy
Audiences today face a multitude of sometimes-conflicting problems. They may face an avalanche of product reviews about certain products such as the latest smartphone offerings. They may have trouble comparing the usefulness and quality of different reviews of a product. They often have trouble comparing two or more different products from different vendors, since reviews tend to focus on individual products. They may be overwhelmed by the choice of products available, but find only some of these products are reviewed at all, and those that are reviewed have a paucity of information. They may feel overwhelmed by apparent the irrelevance of many reviews. They may feel many reviews seem to be more about the reviewer than the product itself. They may feel scared to buy something when finding dramatic negative reviews, or angry when they didn’t pay attention to a negative review and later had a bad experience. They may read many reviews but feel little wiser because the opinions seem confusing or inconclusive.
The current free-for-all in user reviews doesn’t serve brands well either. They have little insight into how reviews are used. They often don’t have a sound operational perspective on how to act on review information. Some brands treat reviews as a social media channel and try to respond to comments as if they were on Facebook. Other brands discourage reviews by asking customers to fill out private surveys. Many Amazon vendors aggressively solicit reviews and even suggest what rating should be offered. Those customers who do submit online reviews may not be representative of all customers.
Brands need to ask themselves some core questions.
How do we know which reviews are influential? When ratings are aggregated, all reviews are considered equal. But not all are equally useful to audiences. Which specific reviews are useful and how do they influence others? Some sites allow readers to rate reviews that are useful. But often “useful” reviews are ones that are long, with lots of description of the product, offering information that should be in the product information, instead of offering true evaluations.
How do we improve the review experience? It can be hard for readers to find the kind of information they seek or that reflect aspects they consider important. They also rarely are able to discover other products they don’t already know about through reviews.
How do we encourage high quality reviews and feedback? Too many reviews are of poor quality, or lacking essential information. Many products aren’t reviewed at all.
Managing reviews as strategic content
Reviews are too important to be left to a junior forms designer on the UX team. Reviews can be a strategic asset, if the right structure is in place to ensure that the content offers value. When viewed through the lens of content strategy, there are many things that can be done to make reviews more effective. I’ll share various ideas on how to improve the product review experience. Not all of these ideas are appropriate for all products, and will depend on the breadth, depth and diversity of products being reviewed. The ideas share a common theme: enrich the information by providing more explicit connections between information items.
Better product review information
Customers often comment on specific product features. They deserve a better structure to enable them to do this. Today reviewers are generally invited to leave comments in one big unstructured text box. This needs to change. One possibility would be to adopt the annotation functionality appearing on some blogging platforms. Reviewers could provide their own comments about their experiences next to the product information describing a feature or product attribute.
Some sites elicit ratings about key product attributes, which allows these more specific ratings to be aggregated. Potentially such attribute ratings could be compared across different products, although in practice this is rarely done. Readers want to know how the experience of customers of one product compares with the experience of customers of another similar product, but online product reviews rarely provide this ability. Part of the problem is that many product-oriented sites lack robust metadata to enable product comparisons; instead, they rely on highlighting what other products people looked at, which may not in fact be comparable. Sites should combine a detailed taxonomy with database queries of product attributes and performance, to suggest what other products are most similar. Ideally, this comparative product set could identify what attributes of are most valued, and which are the biggest concerns to buyers (durability, portability, ease of learning, etc.) Even if these attributes are not explicitly captured, they could be inferred through text analytics.
Finally, retailers and other providers of product review content need to be more proactive in managing the product architecture information. There are so many similar products: variations for different markets or buyer segments, minor product changes, white label products sold under multiple brand names. Using the product taxonomy and product attributes and performance database, the review provider should identify similar product models that have a common basis, where possible. Since consumer review information may be scant for a specific model, it is beneficial to highlight review information about related product models, including information about the brand’s overall reputation.
Providing context about reviewers
The use of real names is not always necessary and may not always be desirable. But providing more context about a reviewer can help readers, and if done properly, benefits reviewers as well. Many sites provide little incentive to reviewers to post their comments, and those that do often reward activity over quality. Sites may acknowledge someone as a “top reviewer” because of the number of reviews posted, regardless of how relevant or useful they are. Review providers need to move away from a social media popularity mindset, and instead think about review posting more in terms of community discussions.
The reputation of a reviewer rests on what they know (post on) and how useful their comments are to others. The concept of reputation points used in community forums can be applied to product review forums. If a reader deems a review useful or not useful, that reputation carries to the reviewer. Reviewers who earn a threshold number of reputation points may receive a benefit that is unrelated to what they are reviewing: perhaps a small sum that can be donated to a cause they support. The recognition of reputation is important for encouraging quality reviews, and helps readers evaluate reviews as well.
Readers also want to know what reviewers know about. By correlating a person’s reviews with product categories, it is possible to provide a high level summary of what the reviewer has written about previously. Readers could see the products a reviewer most actively reviews by brand or category. This scent may offer the reader an indication of other content written by the reviewer they may be interested in seeing. If the reviews are published in a community of interest focused on a product category for fans or enthusiasts, it may even make sense to allow readers to follow a reviewer who has deep expertise in a given area.
In summary, providing more context about reviewers can improve the reader’s evaluation and discovery of relevant products.
Supporting reader needs and actions
Product review content shouldn’t be considered in isolation from the product information: both influence buying decisions. Ideally, review platforms should provide the ability for buyers to choose what criteria is important to them, and see both product information and related user comments about these attributes. With such an approach, it may be possible to use analytics to correlate what review content a reader accessed, with the outcome of their purchase decision. While there are many variables to track (potentially over multiple sessions), if the datasets are sufficiently large it might be possible to infer patterns. It may then be possible to infer the impact of specific review content. What kinds of comments, about what aspects of a product, had the biggest impact on purchases?
Conclusions
Product review content is a strategic asset, and needs to be managed as such. Brands need to move beyond thinking about reviews as a simple popularity contest involving the awarding of stars. Customer reviews are not just another passive KPI metric like customer satisfaction surveys: they are active content that drive customer behavior and business outcomes. To leverage the power of this content, the content must be structured and enriched to support customer goals.
— Michael Andrews