Over the past decade and with the rise of eCommerce, it’s become clear that reviews & ratings data is highly unique as a data source. What’s less obvious is that when businesses utilize them in the right way, they can quickly answer extremely critical business questions like never before.
In a recent webinar with Insights Association, software engineer and reviews & ratings analytics expert, Gautam Kanumuru (Co-Founder and CEO of Yogi), discussed the value and influence that reviews & ratings analytics can have on an organization. The 45-minute webinar also includes four examples of research quarries from real brands to provide an in-depth look into using reviews & ratings data to answer critical business questions.
The recorded webinar is available to view below, and for those who prefer to read we’ve summarized key takeaways in this article.
To paint a full picture of what’s possible using reviews & ratings data, we’ll start by unpacking how reviews & ratings are important as a data source and how we’re able to extract insights from them. Then we’ll dive into specific examples of questions that we can answer using reviews & ratings, highlighting topics such as measuring consumer satisfaction, benchmarking against competition, finding opportunities for product improvement and innovation, and catching consumer trends early.
Why Reviews & Ratings
Why are reviews & ratings so powerful as a data source? There are a variety of reasons for its increasing value and popularity, but for now, we will focus on the top two.
The peer influence that reviews & ratings have on consumers is, in many ways, comparable to word-of-mouth feedback, which has always been one of the most effective forms of marketing. For example, if a friend says that a product is great, most people are far more likely to purchase that product or brand regardless of any advertisements they have seen. In effect, reviews & ratings represent an infinite source of this word of mouth feedback, and downstream, the consumer behavior data reflects this parallel.
This concept was summarized eloquently in a recent article published by McKinsey which provides quantifiable data on the influence that reviews & ratings have on customer decision making.
The article also cites an 80% growth in volume for reviews & ratings just in the past 12 months. At the same time, 93% of consumers read reviews before making purchasing decisions. Additionally, at the time of this research, in this omnichannel, online data-driven world, 40% of consumers are willing to try a new brand or retailer, and that amount has likely increased or doubled since the start of the pandemic. Needless to say, a brand’s reviews & ratings data is extremely important and useful (McKinsey).
However, the influence that the data has on consumer behavior is only one side of the story. The other part of the story is what we’re able to extract from it and why we’re able to find useful insights from the reviews & ratings dataset.
As we alluded to previously, organizations are now realizing that reviews & ratings data represents a real-time, organic, unbiased focus group of tens of thousands of actual purchasers. When an organization can analyze this data with a level of granularity, the insights that they’re able to access become applicable across the organization. This is why so many brands from enterprise-level to up-and-coming have prioritized reviews & ratings as a primary data source for extracting insights.
So to answer the question: Why are reviews & ratings so powerful as a data source? It’s a two-part approach. First is the influence that this data source has on consumer behavior. The second is what this data source represents and why it’s highly valuable.
Now that we have explained the value of reviews & ratings as a data source, we can dive into the question of how we turn this dataset into actionable insights for organizations of all sizes.
How to Get Valuable Insights from Reviews & Ratings Data
Based on our experience at Yogi, there are three core branches to the process of converting raw reviews & ratings into actionable business insights; aggregate, organize, and analyze.
The first step is to aggregate all of the data from a variety of sources. To have a robust data set, it is essential to consistently pull data from across a broad range of retailers and competitors.
Gathering data from a large variety, if not all of the retailers a product is sold in, is important because consumer behavior shifts across different retailers. For example, someone purchasing batteries at Home Depot has different priorities than someone purchasing batteries at Walmart or Amazon. To build a contextual understanding of the market as a whole and the competitive landscape, it becomes essential to have a fully aggregated and holistic dataset.
Aggregating reviews & ratings is the first step to turning it into a robust insights database. The second step of the process is to organize the data strategically or, simply, to make sense of it.
For organizations trying to build a robust system for analyzing reviews & ratings, this is where natural language processing (NLP) and AI-based algorithms become important. It’s one thing to be able to recognize a trend in the star rating of a product, but it’s another to be able to recognize the exact reason and theme that’s causing a trend and how it compares to the overall competitive landscape.
With reviews & ratings data organized properly using the right set of technologies and tools, organizations have access to a much more in-depth understanding of consumer sentiment and can stay agile by making tangible business decisions much more easily.
How do we pull these insights and answers from this robust data source? This naturally feeds into the third piece, which is to analyze the data. After we’ve aggregated and organized the reviews & ratings data set, it becomes essential to be able to query and find valuable insights using a variety of filters and views. This is also where integrating with typical business dashboards or BI platforms such as Tableau or Power BI can become helpful.
As organizations research how to build their engine for reviews & ratings analysis, these three foundational pillars are important to keep in mind. Now that we have defined the value of customer reviews analysis, we can dive into examples of specific questions you can answer using reviews & ratings.
Granular Visibility into Consumer Sentiment
While Rating refers to the star rating that a consumer gives within a review, Sentiment is a measurement that is calculated using an AI system based on the text of the review. In other words, consumer Sentiment is a measure of consumer satisfaction that involves having a contextual understanding of any positive, neutral, or negative mentions within the text of a review regardless of the star rating.
This is important because it allows brands to catch negative mentions within positive reviews as well as positive mentions within negative reviews. Sentiment analysis provides an extra layer of granularity, which becomes extremely essential to analyzing and accounting for overall trends and themes.
For example, a five-star review for a healthy dessert product may say:
“I bought this product to satisfy my sweet tooth while I’m on a diet. The taste could be better, but it gets the job done.”
Surely, if a competitive product enters the market with all the same health benefits and a better taste, this customer will switch brands in an instant. Even though the customer gave a five-star rating, there is a negative to neutral sentiment around taste. By having a Sentiment layer on top of Rating, we’re able to identify product attributes that customers are unsatisfied with.
It’s simple to identify what is happening within ratings, for example, a downward trend in ratings or sales. However, when an organization can granularly break down what’s happening within the text at a holistic level, it can very quickly understand the answer to why those changes are happening and identify a roadmap to solve the root problem.
One of the great things about user-generated content (UGC) such as reviews & ratings is that it’s publicly available. This means that brands can access all of the same data for their competitors, and use that to analyze their performance in context.
On a high level, we’re able to rank brands based on sentiment and rating to quickly give an overall view of where each brand or product fits into the competitive landscape. Within Yogi, we use a bubble chart to provide a view into the big picture of where brands or products rank according to each other, with Ratings on the Y-axis, Sentiment on the X-axis, and the number of reviews represented by bubble size.
By diving even deeper into competitive comparison, we can quickly pick up insights that can drive key business decisions. When we continue to analyze Themes and Sentiment and look at them from a dimension of brand, we’re able to see different elements of the story.
For instance, a product that is underperforming in the overall market may be performing above average in the Theme of Price and Value or Functionality. In this case, the marketing team can adjust their marketing claims and campaigns to lean into this competitive advantage.
On the other hand, the same product may be lagging behind the competition in terms of Durability. Quarrying deeper into verbatim sentences around this Theme and Sentiment, we can quickly identify a specific culprit, for example, a part that is frequently faulty. Now, this becomes an actionable item for the product team. Perhaps they can make a short-term change such as including replacement parts into the overall product offering, while long term they can iterate on this feature.
Overall, by also analyzing the reviews & ratings of competitors, we are again able to generate insights that provide various departments with action items that enable them to make data-driven changes.
Product Improvement and Innovation
What about other forms of product improvement and innovation? By diving into Theme and Sentiment regarding specific product attributes or features, we’re able to isolate consumers’ opinions, likes, and dislikes, regarding that attribute. Furthermore, within those reviews, we can pinpoint consumer requests or preferences to determine the best opportunities to improve consumer sentiment in that category. Simply put, we’re able to understand what consumers are thinking and what they want.
In this case, a product with a variety of flavors may be able to see that Sentiment for one flavor has been falling, while another flavor has been rising, purely in the context of taste. This information on its own is useful across multiple departments, but we can still reach another level of granularity. Within these reviews, using keywords to isolate customer requests, we can identify common customer asks such as the addition or removal of a particular ingredient.
Using these micro-level insights in the context of the bigger picture, businesses can identify minute and yet tangible changes that move the needle in overall consumer Sentiment. As many organizations shift to being more customer-centric and agile, according to McKinsey, brands risk lagging behind the competition if customer reviews aren’t prioritized. Overall, the ability to understand that customers are telling you what they want, what they care about, and what they like and dislike, is becoming increasingly critical for growth and success.
Identifying Customer Trends
With consumer preferences constantly changing, the ability to catch a wave in customer trends early can be the difference between gaining or losing tens of millions of dollars in revenue. This is especially true in the age of eCommerce where the lead time for changing a product description page (PDP) is a matter of days or weeks as opposed to months or more of lead time for packaging changes in the pre-eCommerce era.
Monitoring direct competitors for changes and trends is imperative for any product or brand, but there’s another piece of the equation as well.
For example, let’s say a personal care product can see an increasingly negative Sentiment around Sensitivity. If a new product comes to market that performs better in terms of Sensitivity, this poses a potentially costly problem. Faced with this situation, we can return to the core question; why is this happening? By quarrying deeper into the review verbatims, we’re able to identify the PDP messaging or marketing claim that’s creating a mismatch in expectations in terms of Sensitivity. From there, the marketing or eCommerce team can get ahead of the problem before it arises.
In the new dynamic of eCommerce and innovation, the ability to stay agile and data-driven enough to catch trends early on is essential. Staying ahead of customer trends allows brands to maintain a competitive advantage, whether it’s through updating marketing claims and PDPs or identifying whitespace for product innovation.
With reviews & ratings now representing an unfiltered focus group of thousands of actual purchasers, brands stand to gain unprecedented insight into their audience when they utilize the raw data correctly. To summarize, there are three key steps to unlock the actionable insights that this data source has to offer.
Aggregating the data, or compiling all the UGC available across retailers and competitors, is the first component of a strong foundation. This also includes promotional reviews and syndicated reviews. Followed by organizing it to look deeper into why things are happening. This extends beyond average star rating and review volume by adding layers like Sentiment Analysis and Theme Mapping. Finally, once these two steps are done well, analyzing and manipulating the data allows brands to extract the granular insights that provide the answers they’re seeking within just a few clicks.
As we’ve seen, proper review & ratings analytics allows brands to gain granular visibility into consumer Sentiment, understand the competitive landscape, identify opportunities for product improvement and innovation, and identify consumer trends early on. Using this data source, brands open up to a variety of cross-functional and critical insights that have a lasting impact on their business.
Have you ever tried analyzing your reviews & ratings data? Yogi is designed to help people who aren’t analytics experts get the insights they need to make data-driven decisions fast. If that sounds like you, book a demo today.