Five-star growth: Using online ratings to design better products

New research shows that small changes in star ratings can drive explosive growth for products—on the order of 30 to 200 percent depending on category. Driving those changes at scale requires a novel approach.

As many countries begin to emerge from the COVID-19 pandemic, pent-up consumer demand appears to be driving a strong recovery. But businesses are now operating in a world forever changed. When stores closed and government restrictions forced consumers to remain at home, e-commerce, which had already been expanding, experienced ten years’ growth during the first quarter of 2020 in the United States. One outcome, which may get overlooked amid the current turmoil, is the surge in volume of online ratings and reviews, which were 40 to 80 percent higher during the core pandemic months in 2020, compared with 2019. While this feedback has been available for years, the boost in e-commerce has taken its volume to new heights, giving it greater weight and credibility.

As the number of star ratings and reviews climbs further, and as more products are increasingly scrutinized online, buyers may base their purchase decisions increasingly on cues from other consumers rather than information from companies. If a poorly designed product has a one-star rating and inspires scathing critiques from disappointed buyers, even the most exhaustive, well-funded marketing campaign won’t improve sales. The old levers of brand equity, greater ad spending, or big promotions are simply less relevant in an age when consumers have almost perfect, painless access to infinite word-of-mouth feedback. The products themselves—both the quality and the associated customer experience—may become the most important marketing tool.

With inflation beginning to rise, retailers and consumer-packaged-goods (CPG) companies may need to raise prices to cover their costs. Simultaneously, they face increased competition from start-ups and from companies with relatively inexpensive offerings. But, if consumer reviews of a company’s products are less positive than those of competitors, or even simply on par with them, price increases are not a likely option. In this environment, retailers and CPG companies must give greater attention to star ratings and online buzz as they build a postpandemic strategy. 1 To assist them, we first examined how the COVID-19 pandemic has increased both the volume and the importance of customer assessments. We then quantified the impact of star ratings on sales by product category. Our analysis demonstrates that improvements in a product’s star rating—even an increase as small as 0.2 stars—can deliver meaningful growth in many categories. Our analysis also suggests that insights mined from online reviews are more critical than ever. The same information that gives consumers the power to make better choices also gives companies the power to pinpoint ways to improve products.

A changing market, with digital channels in the ascent

Brand loyalty has been decreasing for the past ten to 15 years, and the pandemic has accelerated this shift. 2 Consumers are now more willing than ever to try new brands, including those recommended on popular sites. Of the 75 percent of Americans who have changed their shopping behavior since the pandemic began, around 40 percent said they have switched brands—more than twice as many as in 2019.

In this fluid landscape, a product’s price and online feedback may be the primary factors that customers consider when making purchases. A generic product that once would have struggled against well-known brands may suddenly become a best seller if it gets rave reviews, especially if it is less expensive than products from established incumbents. Innovative companies often demonstrate the way a product can disrupt a market, even with a relatively small marketing investment. Similar stories abound, albeit on a smaller scale, in almost every industry.

Sidebar

Even though some reviews are false and companies have been known to manipulate star ratings, consumers still largely trust these sources. The surge in online reviews—which rose by 87 percent from December 2019 to December 2020—will add to their credibility (Exhibit 1). While the rate of growth of digital channels is slowing somewhat as the pandemic eases, it will remain well above prepandemic levels—and the volume of online feedback will only intensify. (For more information on this shift and other lasting changes, see sidebar, “The new retail and CPG landscape.”)

The COVID-19 pandemic has led to an explosion in the number of product reviews.
We strive to provide individuals with disabilities equal access to our website. If you would like information about this content we will be happy to work with you. Please email us at: McKinsey_Website_Accessibility@mckinsey.com

Higher ratings equal higher sales

Although companies can easily recognize the link between online star ratings and sales, most have not closely examined the incremental growth from each star. Some research has attempted to quantify this relationship, but we wanted to understand the information at a much more detailed level. To do so, we first analyzed reviews and ratings across the 70 highest-selling categories on a major online platform—hundreds of thousands of individual SKUs over a two-year time span.

The correlation between star ratings and product sales was positive in 55 of the 70 categories we examined. 3 Across all categories, 95 percent of units sold had 3.5 or more stars. SKUs with three- or four-star ratings enjoyed sales that were three times higher than those with one-star ratings.

Even a small rise in score, such as an increase from 4.2 to 4.4 stars, often produced a meaningful improvement in sales. Across the 55 categories, the total growth opportunity—the cumulative sales increase that could be attributed to improvements in products’ star ratings—was 37 percent over the span of the growth horizon. This opportunity varied greatly across categories, however, ranging from 168 percent for cat food to 17 percent for coffee makers (Exhibit 2). Such differences could mean many things; for instance, people buying coffee makers may be more loyal or may simply rely less on reviews.

The sales lift from an improved star rating varied by product category.
We strive to provide individuals with disabilities equal access to our website. If you would like information about this content we will be happy to work with you. Please email us at: McKinsey_Website_Accessibility@mckinsey.com

The beginning of the growth horizon—the star rating at which sales begin to rise—also differed among product categories (Exhibit 3). For many of them, such as office desks and workstations, it began at around 3.5 stars. By contrast, the growth horizon of infant and car seats for children began at around 4.4 stars, most likely because few parents would even consider purchasing a product with a rating below this level, and the risks associated with bad choices are higher. Similarly, the point of diminishing returns varies. Consumers might be willing to pay for the most expensive glucose meter because its safety and accuracy are very important. Buying the wrong pair of socks or paper towels is simply less fraught with consequences.

The starting point for the growth horizon can vary, as can the point of diminishing returns.
We strive to provide individuals with disabilities equal access to our website. If you would like information about this content we will be happy to work with you. Please email us at: McKinsey_Website_Accessibility@mckinsey.com

Another finding: the incremental gains were not the same for each ratings jump. Products with an average review that rose to at least four stars—typically considered a good rating—tended to have the greatest benefits. For instance, one product experienced incremental growth of only 4 percent when its rating increased from two to three stars. But when the rating went even higher, from three stars to four, sales rose by 24 percent. The gain in ratings increased both views and conversion rates.

A new approach to managing consumer ratings and reviews

Despite the obvious advantages of improving a product’s online ratings and reviews, few companies have programs dedicated to this effort. Many, for instance, do not fully understand the value of analyzing review text and thus hesitate to invest heavily in an area that seems speculative or unproven. Others are holding back because of the complexities involved, especially since the pandemic is forcing them to focus on more immediate, pressing issues.

For businesses that do want better strategies to manage consumer ratings and reviews, we have identified six steps that can help them overcome hurdles and achieve significant impact (Exhibit 4):

  • Shift resources into improving products and work more closely with suppliers. Companies could consider reallocating portions of their traditional growth investments, such as marketing and promotions, to more direct investments in their products, in hopes of boosting star ratings and online reviews. In some cases, this shift might involve redeploying staff or adding new roles, especially in product design and engineering. If companies accelerate these changes—for instance, by devoting more resources to engineering or using better tools for rapid prototyping and other tasks—they can accelerate the pace of growth. Close partnerships with suppliers can do so as well.
  • Rethink product claims and improve the customer experience. Companies can sometimes boost ratings through relatively simple changes. These could involve revising product descriptions—for instance, calling a salsa mild instead of spicy to reflect its true nature—or improving instructions to make it easier for customers to use or assemble a product.
  • Manage the product’s digital presence. While it’s critical to get the product right, companies will also benefit from creating a strong strategy to manage ratings, reviews, and other elements of the product’s digital presence. The first step involves encouraging feedback, especially for products with a relatively low number of reviews. Companies may want to leverage “seed” review programs and similar vehicles that retailers provide to bolster review counts. Second, companies should ensure they are fully leveraging enrichment capabilities on retail sites. Providing great pictures, videos, and answers to customer questions can meaningfully increase the conversion rate from click to purchase. Finally, some players have found success by conducting A/B testing to improve their understanding of how search-engine algorithms determine the ranking of search results on retail sites, especially for those with a large volume of sales. Of course, the companies must first consult with the retailers to ensure they are not violating any guidance. Other strategies include increasing spending on up-front advertising to increase clicks, awareness, and initial reviews if the potential return on investment justifies the expenditures.
  • Establish the technological foundation for analyzing reviews. Basic enablers include new data sets and sophisticated natural-language-processing (NLP) tools that can mine free-text comments from web reviews. In many cases, companies will want to expand the use of such tools to other sources, such as social media, YouTube, video sites (by converting voice to text and then feeding it through the NLP engine), industry-specific chat rooms, return logs, customer-care call logs, and even legal claims. Thanks to recent technological advances, NLP tools now make it easier to identify positive, neutral, and negative sentiment in reviews. They are also better at classifying product attributes and picking up idiosyncrasies, such as the use of jargon, slang, or sarcasm in reviews.
  • Adopt new ways of working. Retailers and CPG companies shifting to new tools might need to place a greater emphasis on cross-functional collaboration and organization-wide goals. Engineers will not be tasked only with reducing costs, as they usually are, nor will marketing teams focus only on driving growth through better campaigns. Instead, a cross-functional team will concentrate on improving a product’s appeal to customers.
  • Scale the impact of this work by applying insights across the organization. This scaling process involves some pick-and-shovel work. Companies first take a top-down approach by leveraging advanced tools to pinpoint and prioritize the biggest opportunities across the full portfolio. They then follow a bottom-up, wave-based approach to improve products in different categories, typically focusing first on the most critical ones. The bottom-up work involves methodically improving the designs and overall functionality of each key SKU. Teams should prioritize pain points by using a mix of reviews, social-media comments, and other sources and then identify opportunities to improve margins by incrementally applying design-to-value principles.
A six-part approach can help companies manage online ratings and reviews.
We strive to provide individuals with disabilities equal access to our website. If you would like information about this content we will be happy to work with you. Please email us at: McKinsey_Website_Accessibility@mckinsey.com

Of course, reviews and star ratings are only one part of the product-improvement tool kit. They can paint a picture of major flaws and benefits, but they will not reveal emerging needs or potential problems down the line. In addition, reviews and ratings may not uncover more abstract insights (such as underlying behavioral motivations) to the same extent as traditional techniques such as in-home interviews, shop-alongs, focus groups, and online surveys do. Ratings and reviews will be an important supplement rather than a panacea.

The bottom line: Reimagining product development

While companies cannot control online reviews, they can manage and learn from them. If product teams track star ratings and mine review text for insights, they could create better offerings that reverse market-share erosion, increase the willingness to pay, or even unlock new product categories. In the best cases, better products might help companies leapfrog their competitors by driving much greater organic growth.

One manufacturer, for example, analyzed online customer reviews to identify pain points for a lighting product. It quickly became clear that customers were frustrated because it was difficult to install the light and adjust the sensors, resulting in limited reliability. Guided by these findings, engineering teams redesigned the product to address the issues. With the upgrades, the average customer review increased from 2.5 to 4.5 stars, and sales grew by double digits.

Ideally, engineers will focus on consumer insights as soon as the first reviews and ratings get posted. Does a product feature seem to incite almost universal wrath or lengthy complaints? Do the most positive reviews tend to dwell on certain elements? If engineers understand the customer’s viewpoint, they can set priorities and make corrections before they even touch the next iteration of the product. The time savings from NLP insights alone can be enormous: for instance, engineers may need to spend only 15 to 30 minutes reviewing an NLP tool’s output for a small-volume product rather than fill multiple hours building a design-failure-mode-and-effects analysis (DFMEA) from scratch. With such speed, companies can improve the entire portfolio more rapidly.


As consumer behavior radically shifts in the wake of COVID-19, companies that do not keep pace will find themselves sidelined, especially as more competitors enter the online realm. While retailers and CPG players now have multiple priorities, they cannot overlook how the growth of online reviews and ratings will change their businesses—not just how they market products but also how they develop and redesign them. If they mine consumer feedback successfully and gather detailed insights, their new products will garner even better feedback than their past winners did. This virtuous cycle can continue indefinitely, and companies can achieve breakout growth of 5, 10, or 20 percent in a world where successful growth is usually measured in basis points.

Related Articles