Why business must heed customer reviews

McKinsey research has shown that peer purchasing insights seem to have more influence on consumers than any marketing strategy. In this episode of The McKinsey Podcast, Roberta Fusaro speaks with McKinsey partner Dave Fedewa and McKinsey senior expert Chauncey Holder to understand how and why businesses must prioritize customer reviews—or risk falling behind agile competitors. After, meet Princess Daisy Akita, a mathematician, lawyer, and currently an early-tenure consultant at McKinsey. The following transcript is edited for clarity.

The McKinsey Podcast is hosted by Roberta Fusaro and Lucia Rahilly.

Why customer reviews matter now more than ever

Roberta Fusaro: Chauncey, Dave, thanks so much for joining us today. Online ratings have been with us for a long while, but why are they more important now than ever before?

Chauncey Holder: COVID-19 is the short answer. COVID-19 essentially forced the trial of a number of new categories, and consumers had no choice in the early part of the pandemic but to trial e-commerce as a channel. And surprisingly—or not—they liked it.

When consumers have essentially transparent information about the product that they’re purchasing, it gives them permission to trial as much as they want across the category, right? You don’t have to rely on brand. And so really what you do is you get similar products at roughly the same price point, and now it’s just a matter of ratings and reviews.

And some of our proprietary research has suggested as much—that the increases in e-commerce penetration due to COVID-19 are not likely to recede or regress to prepandemic levels, which means that particular behavior has been adopted widely by consumers.

Dave Fedewa: I think the total number of global reviews something like doubled in the year after COVID-19 started. And so when you think there have been ten years of reviews before that, that’s a lot of reviews in a very short period.

I think that pace may have slowed some, but it hasn’t gone backward. I think everybody’s realized this is an easier, more convenient, and, frankly, more transparent way to shop, and I’m not going back.

The second really interesting thing, which is frequently overlooked, is how reviews affect offline shopping at a traditional brick and mortar. And I think that has also increased substantially. It’s a little bit harder to measure. But just think about the last few big purchases you made. You started that process online. You may have ended up in a store to finish it, but you started that process online. And probably, if you’re like me, you’ve been in a store and flipping through your phone recently, right?

Roberta Fusaro: Exactly.

Dave Fedewa: Like, I’m not buying this thing until I read a few reviews.

Traditional marketing is a thing of the past

Roberta Fusaro: I’m curious, then, what would you say affects consumer’s purchasing decisions more, these online reviews or companies’ marketing campaigns?

Chauncey Holder: We are in a fundamental paradigm shift. But I’m not sure how much practitioners appreciate the tectonic shifts that we are going through at the moment. So, if you think about it, the pandemic has accelerated a fundamental change in consumer behavior.

Going solely into the store and observing what’s on the shelf or any traditional marketing lever—such as the power of the brand, the availability in the store, or some sort of in-store promotion—that’s the old world. If you’re a consumer and your choice is to either put more weight on a signal that the marketer is giving you or more weight on the signals that actual consumers who are using the products are giving you, well, which one do we think most consumers will eventually default to? And we are absolutely seeing that.

Dave Fedewa: That’s a great question. I don’t know if that’s one you could ever answer quantitatively, at least at a macro level. But I think that, for sure, as Chauncey said, the product is becoming the marketing.

In the world before reviews, you needed the marketing to tell you what the product was like, and you needed the brand to give you confidence that “I know the other things that I bought that had this name on it were good, and so this next one should be good.” And quality was a much longer play, longer game, and longer bet. Brand building was a very long game.

And what you’re seeing as a result now is new brands disrupting spaces that had been held by longstanding brands—as this new kind of dynamic plays very well to the disruptor in that sense. The other thing to say, even though it’s a little bit crass [laughs], is you can’t polish a turd. We came across an example recently where a company was paying a reasonable amount for the online promotion of a product. And when you click through the product, it’s rated 3.8 stars.

There’s no point marketing a product that isn’t good, or there’s less and less a point in doing so, because you can’t hide it. So, in a way, you’ve got to fix the product regardless. I’m sure there are lots of different ways to get awareness out there. But it’s absolutely table stakes now to have a great product.

Customer reviews influence consumers to try competing products

Roberta Fusaro: Are there particular product categories where this dynamic is particularly taking hold? For example, I feel really connected to a certain product versus the toilet paper I buy. Is that what’s going on here?

Chauncey Holder: Absolutely. The way that I’m currently thinking about what categories are more susceptible to these disruptions is by thinking about the category along a couple of dimensions.

The first dimension is what is the fundamental benefit or equity that this product is delivering to the consumer? Is it essentially just a functional attribute? Take, for example, a storage container. You know it does a job, and it needs to do the job well. But that’s really the only benefit that it’s delivering. At the other end of that spectrum, you have product categories that deliver some sort of emotional and tangible equity—for instance, product categories for newborns. And it stands to reason that the amount of research and investment that people would do at the end of those spectrums would differ wildly.

The other lens that I think about is in the competitive landscape. Maybe 15 years ago, if you wanted to build a new product and sell it in the US market, one of your biggest challenges would be how do I get inside large brick-and-mortar retailers like Walmart and Target or Home Depot and Lowe’s.

That type of barrier is nil today. So, any enterprising overseas entrepreneur who has some manufacturing and design capabilities can purchase products, tear them down and reverse engineer them, and essentially knock them off and put them online faster than ever—and with very few barriers.

And now you pair that with a post-COVID-19 purchasing population that is more willing than ever to try new things, and then you find yourself with a whole bunch of product categories that are highly susceptible to being disrupted in this new environment.

We did a study where we were looking at the landscape online for a particular product. And we came across an insight that we were not looking for and had not expected. Essentially, in the past two or three years, the number of SKUs available for this product on Amazon quintupled. There was a four- or fivefold increase in the competition online. And what was even more breathtaking about that was the client didn’t know.

Three compounding competitive factors

Dave Fedewa: I think the other compounding factor in all this is that the supply chain is really hard right now, and there are a lot of products out of stock. And so one of the biggest drivers for switching is an out-of-stock product.

I need product X, the one I usually buy isn’t there, what else is here? I’m going to try something else because I need it today. And so you put all that together with great reviews and everything else, and it’s all up for grabs.

To the question, one macrolevel view of which categories this matters the most for obviously depends on price point. If I want to spend a lot of money, I’m going to research it, right?

The second factor is emotion, to Chauncey’s point. Take, for example, baby car seats. We conducted some research: baby car seats are one of the most sensitive things when comparing the relationship between growth and star rating. You can just imagine the conversation: “Honey, I think the 3.8-star car seat should be fine, and it’s $50 cheaper. Are you cool with that?” And just imagine the response you’re going to get. Now, it depends on whether it’s your first kid or your third kid probably a little bit, but—

Roberta Fusaro: Exactly.

Dave Fedewa: The third factor I’d say is that even if it’s not expensive or something I’m emotionally wrapped up in, how much of my life is going to be tied to this? And we looked at this in home improvement, and some of the ties came out surprising. Things such as deck screws or a low-cost toilet valve or something, right? And it’s like, holy cow, it matters a lot. The star rating really matters a lot in these categories. And for a minute there, I thought, “Is this right?”

But then what you start to realize is that someone’s going spend an hour or two of their Saturday, or maybe a lot more, and yeah, it’s only $5 or $10 or whatever, but they care that they don’t want to do it again for a long time. And so, even in lower-price-point items, it matters.

How best to leverage tech to analyze reviews

Roberta Fusaro: You wrote an article recently for McKinsey on how to take online reviews in hand and figure out how to improve the design of your products as a result. What sort of base technical capability do you need to establish to analyze reviews to get the insights that you need to make these differences in your product categories?

Chauncey Holder: The good news is that it’s neither very complicated nor very sophisticated. Primarily it’s called natural-language processing, or NLP—and that’s using machine learning and, maybe to a lesser extent, artificial intelligence. We crawl websites so that we harvest the consumer reviews, which are publicly available.

And then the natural-language processing takes these reviews—which, in the industry, are called unstructured text—and it does two things. It clusters the reviews into themes, and those themes are usually very product-attribute related because of the nature of the reviews. And then, it can take that large data set of unstructured text and cluster it into product attributes, and it can assign a positive, negative, or neutral valence to a given review.

And the output of that is an attribute taxonomy that is essentially from the mouths of the users, and a performance metric, which is your ratio of positive to negative reviews for a given SKU or brand on a given product attribute. So you get a highly actionable diagnostic tool simply by using natural-language processing.

Dave Fedewa: The secret to having a successful product business hasn’t changed for hundreds of years. It’s iteration, right? Get a product out there, and you obviously start with prototypes and what’s not working. You fix that. But then once you get it out in the market, you’ve got to continue that, right?

And I think you could take every successful company and say that they’re doing a good job with that. You could take a lot of unsuccessful companies and say they lost that somewhere. They didn’t keep improving, and they lost their edge. Elon Musk actually captured this really well recently in an interview with the Wall Street Journal.

They were asking him about innovation, and he was in front of a bunch of CEOs. And they asked, “What would you tell this group about innovation?” And his answer was along the lines of, “I would tell them it’s not actually that much about innovation. Instead, I would say it’s about asking yourself is my product as awesome as it could be? And the answer to that’s probably no.”

It’s really easy to go out and find out what’s wrong with your product—what people don’t like about your product and how you compare by attribute to your competitors. And most importantly, what are the most important attributes? What do people care the most about? And then conduct an analysis of the things that people are least happy with about the product to help prioritize what you need to fix first.

The next problem with that is, OK, I know what I need to fix, but it costs money, right?

Now we can be reasonably scientific about that. And so the beauty of that is now I can build a business case for the cost required to fix the top three issues with my product and be pretty confident about the corresponding growth that’s going to come from that fix.

Iterate quickly

Chauncey Holder: I would just say on top of all that, the dynamic in the marketplace is such that the length of time that you have to do those iterations is smaller than it’s ever been and will only continue to decrease because of the ease of entry for competitive products.

And that’s the bad news. The good news is that ratings and reviews as a source of consumer insight are nearly real time and—relative to other traditional marketing research techniques—very, very fast and very inexpensive.

Roberta Fusaro: Can either or both of you provide examples of companies that have used these online ratings as an asset and took input from consumers and somehow improved their product?

Dave Fedewa: One example in home improvement is outdoor lighting. There was a product where part of the value proposition of it is that even when you walk up to it or if it senses any motion, it kicks on, or it gets brighter. And there was an issue with that product where that wasn’t happening on a consistent basis.

As you can imagine, if somebody spends a few hours on a Saturday putting one of those up, and then they bring their family and say, “Hey, watch this thing work,” and then they walk up to it and it doesn’t work, they’re not going to be happy. In that particular case, the company went through the effort to make it better and redesign it. And they saw substantial growth in sales as a result.

Chauncey Holder: I think we have a case study on this brand that originated as a generic brand selling flashlights and software peripheries—such as cables and mice—primarily on Amazon. And they were really pioneering in how to come in and disrupt the marketplace and how to meet consumers on their own terms and figure out how to leverage these dynamics. They were one of the first companies or brands that would mine reviews extensively on their products, not only to make product improvements, but also to quickly get out in front of dissatisfied customers who were leaving reviews and course correct that.

One of the things that we’re seeing is that successful entrants in these online marketplaces are the ones who are the most responsive to the reviews, who are most proactive in soliciting reviews, and who are doing the best job at what I would call online-marketplace hygienics. And we’re seeing that many incumbent brands are not keeping up with the basic hygienics of how to market your products in these marketplaces.

But this particular example has done so well online and at Amazon that you can also find them in some brick and mortars now. And it all is because they were taking the feedback that consumers were giving them and improving their products iteratively over time. Simple as that.

Marketing must be accurate

Roberta Fusaro: I’m wondering if you could talk a little bit about what companies can do to make sure that consumers’ expectations match the actual product that they’re buying. How can companies clean up their language to create that dynamic?

Dave Fedewa: So first, be clear about your product and what it is—and in some cases what it isn’t. We were looking at a brand of salsa the other day that proclaimed to be a spicy salsa. When we went through the reviews, it was very clear it was not spicy; it was mild.

So that’s easy to fix: just call it the mild salsa, right? Or you can reformulate and make it spicy. But call it what it is so you don’t disappoint. There’s not a lot of science behind that. And now it’s very easy to find those cases and address them. It happens all the time.

Second, be aware not only of the product and its performance but also of what the experience is going to be like and what it’s going to take—with home improvement, for example, that could be ease of installation or instructions. The two words, “No tools,” are really powerful. It’s not necessarily the first thing you’d think to call out on, say, a desk, but to be able to say, “No tool assembly” on a desk is a big deal.

If someone might think they’re going to get something out of a product that they’re not, you’ve got to be very clear about that. Because it’s going to come back. It’s going to return. We haven’t talked about that, but that’s a huge piece of the economics here. Returns are way higher online. So don’t be shy to say, “It’s this and not that. And if you want that, that’s a different thing.” And by the way, that’s probably another SKU you should have. That’s an upsell.

Making the case to put budget into product quality

Roberta Fusaro: Dave, you mentioned how it’s easier, or maybe more common, for companies to make the business case for how to move budget around to prioritize product improvements. I wondered if you could talk a little more about how companies should think about doing that.

Dave Fedewa: I think there are a couple of different levels to it. You start at a macro level and say, “Across my whole portfolio of products, where do I have the biggest star-rating growth opportunity?”

And that’s something again that’s easy to do, right? I can go scrape all of the star ratings for a company’s entire portfolio, and then do the same thing with their competitors, and break it down by major category for the company and say, “OK, in which of these categories are you leading versus lagging behind your competitors?”

And which ones, frankly, are you lagging behind in the most? And then let me also stack that up against how much revenue each of these categories drives. And then use a little bit of the science and research that we’ve done and say, “OK, so what is the percentage growth opportunity by category based on my star-rating gap?” It’s not decimal point accurate, but it’s a really nice way to ask, “What are the first three categories I should probably start using this thinking on where I’m going to get the most growth?”

And you can even size it. You can put a dollar number on it. And we’ve seen dollar numbers in the hundreds of millions because fairly frequently we’ll see percentage increase numbers of 30 to 50 percent—going, for example, from a 4.2 to a 4.5, depending on the category, is worth a lot of growth. So that’s first at the macro level: if, say, this is worth $500 million in growth to my company, then I’m ready to probably dedicate some resources and tools to that growth, right?

The next thing is to determine on which three categories do you use this thinking? Then, we move to the micro: Inside these categories, which SKUs have the biggest opportunity? And then the question becomes, on those SKUs, what are the three or four or five attributes that I need to fix first?

And I can get that too—I take all the negative-sentiment comments and do an analysis. And then maybe issue number one is the instructions. All right, I can fix that. I need better instructions. Let’s get somebody on that. We know that’s worth a lot of money.

Issue number two could be that I’m calling it spicy instead of mild. So let’s go fix that, right? Issue number three might be more complicated. Maybe I’ve got a strength problem, and I need a different alloy. But I can make the business case for that and see that if I fix that problem, here’s what I’m going to get in terms of growth and incremental margin. So I’ve got cost and benefit. They’re very straightforward.

The customer review’s influence is here to stay

Roberta Fusaro: That sounds great; it’s a very systematic approach. Is the growth and volume of online reviews and their increased importance here to stay?

Chauncey Holder: A hundred percent.

Dave Fedewa: No one’s going to trade a more time-consuming, less transparent, in-store-only experience for a less time-consuming, more transparent, online experience.

For a long time, it was, like, well, I’ll do it with toothpaste, but I don’t know about a grill, right? And then, well, I’ll do it with a grill but maybe not a refrigerator. And then, well, I don’t know about a car, right? And now, more and more, it works for everything.

Chauncey Holder: It really does. There was an article in the New York Times recently. There’s an online auction marketplace for vintage cars and enthusiast vehicles. Prepandemic, most of their auctions were $20,000 to $50,000 automobiles.

Postpandemic, they’re able to auction vehicles anywhere from $400,000 to $1.5 million dollars. And it’s completely virtual. People are willing to do a $1.5 million transaction for a product that they will not physically touch or examine. Now, if that doesn’t bring into sharp relief a fundamental shift in consumer behavior across the board, I’m not sure what else would.

Roberta Fusaro: Thank you, Chauncey and Dave, for joining us.

Dave Fedewa: You’re welcome. Super fun conversation.

Chauncey Holder: Been a pleasure.

Segment two: Meet consultant Princess Daisy Akita

Princess Daisy: I’ve been at the firm for almost a year, and it feels like, “Oh my God, how could I have done so many different things in so many different contexts?”

Before McKinsey, I had just finished a JD and taking the bar exam, and I was very much in the space of lawyer thinking: super detail-oriented. Lawyers go into work looking for periods and commas. Coming here, I experienced a slightly different value paradigm, harking to the 80/20 principle, where the first 20 percent of your effort can give you 80 percent of the value.

And that has really reinforced the idea that sometimes, it’s good to refine an answer over and over, but oftentimes, there’s also value in going with your gut, seeing what the data tells you at the very beginning, and seeing what story you can build from that.

I grew up in Ghana. I lived in Ghana until I was about 17 and then moved to the US for the last two years of high school. I studied applied math during my undergrad, and I really liked the very precise, very quantitative nature of that. Fast-forward, a few years later, I was in law school, which is like the exact opposite of that: very meeting heavy, very qualitative.

So I thought it was fantastic that consulting existed—and I could leverage data in a way where storytelling counted—and allowed for the bringing together of those quantitative and qualitative interests. Frankly, I can’t think of many careers where I can build all of that into a single job on a day-to-day basis, and then still get to do tech work, which I think is really interesting. So I think that consulting is great for people who have a ton of interests.

Most of my studies have been in the SHaPE Practice, which is the Social, Healthcare, and Public Entities Practice at McKinsey. There’s a lot under that umbrella—we work with clients in the healthcare space, and also federal, state, and local governments.

My first SHaPE experience was related to COVID-19. We were working with a state government to try to roll out vaccination across the state, which, as you can imagine, was incredibly exciting and incredibly impactful and intense because we needed to make sure our state partners were ready and able to get that vaccination machine up and running.

There were more than 30 people doing work on aspects of the same topic. And if you think about it, these are 30 people who are thinking and breathing the very same problem and making sure that everybody is on the same page and that we are continuously aligned about what the goal is. And this was a rapidly evolving situation, not to use that term loosely, but it was rapidly evolving, and so you need to be educating and re-educating the team. And that absolutely will remain a hallmark of my McKinsey experience.

I do feel that there is a sense of possibility, and there’s a sense that if you can envision something, if you can build a compelling case for it, there’s runway for you to bring that to fruition within the firm. One thing I really appreciate about McKinsey—and this is in contrast to some of the other consulting firms—is how much agency you have over what work you do. You’re able to select which study you’ll end up working on. So for the next three months, you’re thinking about this problem. And if this problem, and this group of people, feels inspiring to you, you get to follow your heart. And you do that over and over again. And, eventually, you build a McKinsey experience, and you build a life that you can appreciate, and you can celebrate in the day-to-day.