Twenty Years (and 10k Cites) of "EWOM"

“EWOM is “any positive or negative statement made by potential, actual, or former customers about a product or company, which is made available to a multitude of people and institutions via the Internet.”
— Hennig-Thurau et al. 2004, p. 39

THE article “Electronic word-of-mouth via consumer-opinion platforms” I had the pleasure to co-author with colleagues and friends Kevin Gwinner, Gianfranco “Johnny” Walsh and Dwayne Gremler, came out in the Spring of 2004 in the Journal of Interactive Marketing. Twenty years after we introduced “EWOM” to marketing science, the article is now considered a seminal piece in the field of digital communication and marketing¹; it has garnered over 10,000 citations from academics inside and outside our scholarly discipline, more than any other not just on EWOM, but also on word of mouth in general.²

Here’s my story of the genesis of our initial EWOM study. Let me take you on a personal journey through time, from EWOM’s initial days via today’s omnipresence of digital customer reviews into the (nearer) future, analyzing what my fellow co-authors and I anticipated back then, but also how the concept has grown in importance and changed since then — and what we did not see coming twenty years ago. I also want use the opportunity to lay out a vision of what I consider the most pressing developments for scholars in the still vibrant but now also often concerning EWOM landscape.

The Early Days: Introducing “EWOM” to the World of Marketing

When I think back about my research on what is now known as electronic word of mouth, or “EWOM”, one dialogue stands out. It must have been sometimes in the spring of 2000, when I revealed my ambition to conduct research on the reviews of goods and services that consumers were beginning to write, and read, on the internet to Ursula Hansen, my academic mentor during those days.

This phenomenon, I hear the younger version of me saying, will be as relevant one day as the product reviews of Stiftung Warentest, the German equivalent of Consumer Reports in the U.S.! I remember that Professor Hansen, at that time a member of the advisory board of the Stiftung, was somewhat skeptical about that young man’s bold claim, to phrase it mildly. But she nevertheless made the connections to those who became sponsors of what turned out to be the first of many collections of EWOM data.

Those were the days. For understanding how, and where, our research journey began, it is essential to understand how different the world of marketing was back then from how we experience it today — for marketing managers, but even more so for consumers. The internet had only set sail a few years before — just one out of six Germans of 14 years and above had access to the digital network in 1999.

The concept of word of mouth had been known by managers and marketing scholars alike for quite some time, but neither paid much attention to it. Without the internet, it was almost impossible to empirically measure personal conversations about brands and products, and word of mouth was structurally inferior to the powerful propaganda tools marketers and salesmen had at hand. To reveal scams, consumers had to rely on the mass media, but journalists were almost impossible to reach back then, when direct links between journalists and consumers were science fiction. (Remember that email was far from being a mass medium, and tweeting was something that birds did, not humans.)

In the analogue pre-digital world, hardly any consumer voices were loud enough to compete with a nationwide ad campaign. I remember these days, which nostalgics now often call the “good ol’ times”, as the age of bad products and service deserts, characterized by little transparency and a general lack of consumer power.

But the internet, and a number of digital entrepreneurs, had just begun to change all that. In the United States, the website Epinions had begun to publish consumer reviews of goods and services in September of 1999, and in the spring of 2000 four of such consumer opinion sites (that was the common label back then), namely “Ciao”, “Dooyoo”, “Hitwin”, and “Vocatus” competed for the attention of consumers in Germany — and consumers’ contributions in terms of written product reviews. All of them were instant successes, with millions of monthly page visits and impressions. 

I sensed that there was the potential for a power shift with EWOM, from the marketer to the consumers: the spread of a new source of information that allowed consumers to raise their voice and disclose marketers’ wrongdoing and tricks. I envisioned a better world for consumers, maybe a new age — one of great, customer-centric products this time! But at the same time, I have to concede that I had no idea of the seismic dimension of the changes that would come along with the rise of EWOM and which would lead not only to the disappearance of the initial opinion platforms (of which only Vocatus still exists, now operating as a consulting firm though) and disrupt how we decide for (or against) a product, but which would also affect the fundamental concepts of truth and democratic values.

Collecting the data. Before I get to what happened to EWOM since then, let me share some memories on the long and quite winding process of writing and publishing what now is considered a seminal article on EWOM, as what happened to us back then might be informative for those who study new phenomena now. First of all, the data collection was quite painful. Keep in mind that those were the early days of the internet — the survey engine that was developed specifically for our project (online market research was very much “under construction”) suffered a number of breakdowns over the Christmas period of 2000.

A key take-away from those first steps into EWOM territory, based on responses from my family: Never ever collect data during the holidays. But we made it in the end, having gathered responses from more than 4,000 consumers, almost 3,000 of whom had read EWOM and about 2,000 who had articulated EWOM themselves. Back in the days, these numbers provided a first glimpse at the relevance of the new phenomenon.

A not exactly euphoric reception. The initial reaction from almost anyone I talked to about the study was skeptical, if there was a reaction at all. Remember those where the days where marketing managers disputed the mere relevance of the internet in general, and the burst of what is now known as the first “internet bubble” of dubious online business models didn’t help much to establish the internet as a serious research topic.

Believe it or not, the internet back then was a nice, funny little thing that many people thought (and hoped) would disappear overnight, just like it had arrived at the scene! Marketing academics back then predominantly considered the internet as a cheap way to collect data from a parallel universe: “You can easily collect big data sets there, but they don’t mean anything in the real world”, I remember a senior colleague saying at a conference. In the case of EWOM, the main challenge was even harder: Not only did the data come from an internet survey, but the core concept itself was inseparable from the internet, as the only place where you could find online reviews and EWOM. So the whole phenomenon was often trivialized and ridiculed.

A pivotal moment was a German marketing conference at Bad Homburg, where about 120 German marketing professors gave me clear advice: If you want to get ahead with your career young man, do something serious, not that “internet stuff”! In hindsight, it seems obvious that these confrontations where, like so many that would follow over the next decades, part of a paradigmatic and often inter-generational battle: The clashing of those who were intrigued by digitalization and the changes it carried with those others who defended the relevance of their analogue knowledge resources (and power base) in terms of both concepts and methods. Quite fiercely they fought (some still do), I have to say, and what may sound like a perfect example of an academic debate actually had more resemblance with hitting and stabbing from time to time.

Assembling the team. But certainly not everyone was critical, and some even shared my enthusiasm. The first scholar who joint me on board of the EWOM train as early as in the spring of 2001 was Gianfranco “Johnny” Walsh, then a striving post-doc at Hannover University (to where he recently returned, now serving as a full professor) — and a good friend until today.

For John and me, getting our findings about EWOM out into the world was part of another distinct struggle, one against the self-sufficient nature of the German academic system, where scholars had often little interest in participating in, and contributing to, the global scientific debate. Note that only a few years before we started our EWOM quest, Hermann Simon, then a marketing professor at Mainz University, had compared the academic system to a black hole — a system that was sucking in knowledge from the “outside”, but did not care to contribute in return. His observation might have contributed to him saying farewell to the academic system (and founding what became a leading global consultancy, Simon-Kucher).

A few months later, American marketing scholars Dwayne Gremler and Kevin Gwinner, who I had got to know as inspired collaborators on another project and who had since become good friends, completed the EWOM team. Dwayne had an additional motivation: Some years ago he wanted to do research on (traditional) word of mouth, but was told by a senior eminence that “WOM is dead”… With them on board the EWOM train, we felt like the four ‘marketing musketeers’ who had assembled, and writing the paper picked up full speed. Not only did Kevin and Dwayne’s involvement elevate the project to the next level; their excitement also gave John and me a foretaste of the global potential of the topic.

Giving up (almost). Over the next couple of months however, my enthusiasm for the topic was somewhat dampened once more as a result of the rather muted responses we got from many peers. When it came to choose a target outlet for the paper, we decided to skip the marketing discipline’s “A+” outlets because of their reputation to be quite skeptical of unfamiliar topics. (Something that I believe hasn’t changed dramatically since then…)

So the Journal of Interactive Marketing, with its curiosity for ‘everything digital,’ was the first outlet we sent our paper to, and that turned out to be a wise decision. JIM’s then-editor Venky Shankar was a big help — not only was he open to the topic, but he also requested us to add the regression analysis that links EWOM motives with EWOM behaviors, over and above the typology of consumer motives (which was the core contribution of the original submission to JIM). Doing so, Venky wrote back then, would “considerably strengthen your paper”, and he could not have been more right. The results revealed that the articulation of EWOM, at least in its early days, was a mix of social (being an active part of a community and concern for those consumers who would read the reviews) and intrinsic (self-expression!) as well as extrinsic incentives (rewards by the platform).

On October 16 2003, JIM’s managing editor dispatched the final acceptance letter, which arrived a few days later in our post boxes (along with the request to send “the disk and hard copy” of the final paper version — a regular part of life before the internet became mainstream).

EWOM awakening. After the publication, reactions to our article, along with those to a complementary publication on EWOM reading motivations that John and I got accepted by the International Journal of Electronic Commerce, remained quite limited over the years that followed. During 2009, five years after publication, there were still “only” new 116 new citations of our JIM article, according to Google Scholar.

It was the exponential rise of the internet, and the widespread use of consumer reviews on shopping, but also video sites (such as YouTube) and social media, that had begun to exert its big impact on consumers’ and managers’ lives. I remember quite vividly the day a radio moderator from the public radio station Bayern 2 contacted me in the summer of 2011 and invited me to participate in an one-hour call-in broadcast about “evaluation platforms on the internet”. They obviously considered me the leading expert on the topic, and as I wanted to participate I didn’t tell them that I hadn’t done anything on the topic for more than seven years, and that the data collection for our study, the source of most of my EWOM insights, took place even almost a decade ago…

Such growing interest in EWOM was not limited to journalists; digitalization also drew marketing scholars’ attention to our seminal article on the EWOM phenomenon. Over the course of 2014, another five years later, now 500 publications cited the article, and in 2019 the number of annual cites had risen to almost 900. In January 2024, the article’s citation count reached the 10,000 mark, coinciding with the 20th anniversary of its publication. (For reference, fewer than 40 other articles from marketing’s Top 20 journals have ever achieved this level of citation.) Today, the article ranks as the most cited not only in the field of EWOM, but also among all research about word of mouth and consumer communication in general, according to the Publish-or-Perish software and literature reviews.

Today: Evolving Toward an EWOM World

Now let me take a look at the many things that have happened since the publication of our manuscript, which allows me to also shed some light on how the concept of EWOM has evolved during its first 20 years of existence. While there are a number of things I believe we did right, other developments were clearly beyond the scope of our vision back then.

What we did right. Even though all of us four authors were more or less rookies to the world of marketing academia, we risked studying something that was new and for which no established theory existed. Why did we do so, despite discouragements from peers and skepticism of our mentors? The answer is simple: Because we were convinced that our observations, substantiated by our very own changes in communication and consumption behaviors, were neither accidental nor outliers, but instead pointed at a substantial shift toward a new digital world of marketing we were fascinated by.

It is widely known that academia, in marketing as well as in many other disciplines, has a strong bias against radically new topics and phenomena, despite its often-declared willingness to embrace them. Our EWOM story, and the mere fact that this essay exists, should serve as an encouragement though for its readers. If you are a young scholar intrigued by something novel that affects consumers and marketers, follow your excitement. The road might be (even) more cumbersome than for incremental research, but there’s a sound chance that the outcome will be worth the pain, on a scholarly level, but also personally, as there is no substitute for studying phenomena that are practically meaningful.

The (many) things we missed. Our analysis of EWOM focused on consumer opinion platforms, which were pretty big at that time and clearly the main resource for EWOM-searching consumers. We even named the phenomenon under scrutiny “customer articulation on opinion platforms”, alias “CAOP,” and changed it to EWOM only five months prior to the initial submission—a decision that prevented opening a different path through time in which this essay might not have existed…

Those platforms, such as ciao.com, dooyoo.de and hitwin.de, had made consumer reviews the core of their respective business models, stimulating EWOM to generate advertising revenues and shares from affiliated retailers, sometimes by offering financial incentives. We cooperated with them and met their CEOs. While they were all bright and brave, it turned out that none of them had a vision for EWOM’s future as powerful as Amazon’s Jeff Bezos, who made EWOM a key asset for his retailing business, creating network effects and benefiting from them hugely. As scholars, we did not envision then that retailers would become the leading source for the kinds of online reviews from consumers we were studying, whereas all of these “original” review platforms went out of business earlier or later.

To our defense, Amazon was tiny compared to the giant it has become since then (the German branch was established only in late 1998). Kudos to Mr. Bezos, whose business legend is closely tied to the publication of consumer reviews, which his firm began offering as early as 1995. What we didn’t see back then is that a commercial company could harvest the power of customer-centric marketing by pushing it to such an extreme level. Mr. Bezos saw customer reviews as a core pillar of his business model.

Back in the days when he was still running a book store, he got letters from publishers saying, “Why do you allow negative reviews on your website? Why don’t you just show the positive reviews? Maybe you don’t understand your business. You make money when you sell things”. His response was inspired by nothing less than the definition of marketing itself — that is, addressing customer needs ultimately leads to success. He prioritized EWOM over pleasing publishers for the same reason: “We don’t make money when we sell things; we make money when we help customers make purchase decisions,” he said in 2007. With Amazon, Mr. Bezos uplifted customer centricity across the entire business world, and the presence of EWOM (both positive and negative) on the retailer’s website was both a reflector and a key element of this strategy.

Other, more specialized “retailers” have since then embraced EMOW in a similar way to Amazon — think of Airbnb and Booking for the accommodations they rent and Netflix for the movies it streams. The same applies to some general information platforms: Google for all kinds of locations, IMDb for all kinds of movies and series (that Amazon owns the latter site should not surprise anyone…). What happened to the original opinion platforms, you may wonder? Some, like dooyoo and ciao, tried to follow suit and transformed themselves into shopping portals. Ironically, they ignored one of the key secrets of EWOM success when doing so, namely that readers’ trust in reviews is essential. As a result, they prioritized sales-boosting tactics over genuine critique and had to succumb to the diminishing credibility that resulted. Some of their successors such as Yelp and TripAdvisor still exist, but they attract clearly less traffic now than they did in their heydays a decade ago.

Another development we did not anticipate was the fragmentation of customer reviews into various kinds of EWOM. When we started our exploration of the phenomenon, there were no social media platforms on which consumers could post their experiences with brands and products — Facebook dates back to 2004, the year our article eventually appeared. Instagram was even founded six years and TikTok full 12 years later… There was also no microblogging, with Twitter being introduced full two years after our article was published.

Together with other scholars, we have since then tried to address this development, based on the conviction that fundamental differences exist between consumer reviews and “social media WOM”, as the latter shares as many similarities with personal communication as with consumer reviews. Think about the personal connection between sender and receiver in social media settings and also the ability to provide almost synchronous feedback to a social media friend’s product-related statements — two issues that do not apply to consumer reviews at Amazon et al. Specifically, Caroline Wiertz, Fabian Feldhaus and I proposed a typology of different kinds of (E)WOM that highlighted the distinctive nature of reviews on websites and articulations on social media platforms. Joined by André Marchand, Caro and I also provided empirical evidence of the unique roles the different kinds of EWOM play for consumers’ decision making, and Maria Bartschat, Gerrit Cziehso and I recently did the same for information search behavior. 

Since our inaugural investigation into the topic, EWOM has also become multimodal, substantially expanding its “text-only” feature from the early days of the internet when infrastructure limitations (it was the age of the modem!) prevented consumers’ from sharing videos and often even photos. (And remember that most of us had no devices at hand to even record videos…) Together with Edward Malthouse and five other colleagues, I made up for this in a follow-up paper six years later, when YouTube had just begun to publish consumer-generated videos (in 2005). We cited Dave Carroll’s “United Breaks Guitars” video as evidence for the enormous impact multimedia EWOM can have on consumers, in a way foreshadowing the title of the singer’s book about his year-long dispute with United Airlines, the company that broke Mr. Carroll’s Taylor guitar: “The Power of One Voice in the Age of Social Media”. Today the role of the respective EWOM mode (text, photo, video) for the spreading and influence of an EWOM post has become a research interest on its own, and rightfully so.

One additional later refinement of the original EWOM concept sets those consumer statements that are based on true experiences with products and brands apart from those which are purely speculative. Personal experiences from fellow consumers have always been the core value source of word of mouth for those who base their decisions on it, and while we did not restrict our original EWOM definition to experiences, the rise of speculation and rumors in digital media has made a conceptual distinction worthwhile, if not essential. Together with my fellow Entertainment Scientist Mark Houston and colleagues Ann-Kristin Kupfer and Martin Spann, we compiled rich qualitative and also quantitative evidence that such buzz deserves to be seen not just as a “buzzword,” but also as a theoretical construct. By defining buzz as the aggregation of observable consumer anticipation about a new product, we disentangle its speculative nature from experience-based WOM (and EWOM). Thus buzz should be considered not as an informal substitute phrase for EWOM, but as a complement to it which has distinct antecedents and also a differential impact on new product success. (The review board of Journal of the Academy of Marketing Science seems to have agreed with this logic when awarding us best article of 2018!)

Most importantly though, we had no idea back then just how BIG EWOM would become. Today it constitutes the most-used information source, with now more than 8 out of 10 consumers consulting EWOM before buying products, even more than those who seek face-to-face WOM, and WAY more than the users of search engines and social media. EWOM is even more popular with Generation Z, where almost everyone makes use of it. In academia, the publication of not one, but two meta-analysis articles in marketing science’s leading journals within recent years provides evidence of EWOM’s prominent role these days. But realizing how dramatic the changes would be that the ever-accelerating digitalization exerts on us took some time.

Once we did, we crafted the “pinball paradigm” of marketing, which has EWOM at its core, as a powerful way to reflect these changes. The pinball paradigm, jointly developed and introduced to marketing academia by Ed Malthouse, Bernd Skiera, Christian Friege, Sonja Gensler, Lara Lobschat, Arvind Rangaswamy and me, retires the “bowling” approach to marketing, where marketers have the power to “hit” consumers through unidirectional mass media messages, just like bowling balls hit pins standing at the end of the bowling alley. Instead, in the new pinball environment everything has become chaotic and probabilistic, with consumers taking marketing stimuli like ads and commenting on them in their very own EWOM way, celebrating, and criticizing, modifying and parodying them multiplying their reach by sharing them virally with unlimited numbers of other consumers, as well as with journalists. With Charlie Hofacker and Björn Bloching, I later elaborated what successful pinball marketing requires, and Mark Houston and I explored the paradigm’s meaning for movies, games, and other “hedonic stuff.”

Occasionally consumers’ actions even create social media firestorms, when the collective online engagement passes critical thresholds and responses are returned with massive force to the sending firm and beyond. They can have quite drastic, and long-term, effects for firms who are hit by such EWOM firestorms, as Nele Hansen, Ann-Kristin Kupfer and I have shown in an effort that was awarded with the best IJRM article prize in 2019. In a certain way, it seems appropriate to argue that the advertising-dominant world of marketing has transformed into an EWOM world. 

The Future: To EWOM Infinity and Beyond!

Twenty years have now passed since the publication of our initial work (a quarter of a century even since the collection of the data set), and I find it stunning to see that the EWOM research landscape appears as thriving as never before. A devastating pandemic has shifted many parts of our lives further toward the digital realm, and so it seems quite certain that EWOM is here to stay. At the same time, I assume that the concept will further change in conjunction with the dynamics of our societies and technologies, as it has done so much in the past.

For providers of goods and services, the rise of EWOM toward a dominant source of consumer information has made the production of high-quality products an essential requirement. The “Twitter effect”, as Caroline, Fabian and I labeled the sharing of quality-related information by consumers almost immediately after they have experienced a new offering such as newly released movie, based on industry verbiage, is a focal part of the pinball environment. It lets information about product flaws travel with enormous speed, and the only way today for marketers to counter such quality-related transparency would be to spread misinformation, something that should be off the table for every company and marketer who agrees with basic ethical values.

But where does such unethical behavior begin, when it comes to managing EWOM though? The stimulation of EWOM by happy customers can be usually considered within the realms of ethical behavior, such as when a firm sends follow-up emails or adds little notes to the packaging that motivate the buyer to post a review, in case he or she likes the purchase, like in the example below.

But certainly limits exist. While asymmetric treatment of happy vs not happy customers might be tolerable, financially incentivizing customers for posting five-star reviews is a more dubious, though common industry practice, as it provides a strong motivation for customers to praise the product regardless of its actual performance — or even buying the product to earn money when the fee exceeds the price of the product. Why would companies do so? Because the value of a positive EWOM review at the beginning of the product’s lifecycle could exceed the losses of early transactions. A related question is how companies should behave if their competitor embraces such unethical tactics.

Probably the biggest challenge for providers of goods and services, however, results from the plethora of different EWOM formats and types. Firms now need to understand which EWOM formats should be monitored to understand consumer responses and to prevent firestorms, along with the formats’ prioritization. Based on this information, they will need to create EWOM dashboards that monitor how their offerings are doing on each of the relevant EWOM formats — and to update this information relentlessly, as new formats and platforms constantly gain importance while others lose relevance.

Among those EWOM platforms that I assume firms need to look out for is the “metaverse”, a concept popularized by Mark Zuckerberg’s renaming of his Facebook company to Meta that describes a computer-generated environment consisting of virtual (3D) “worlds” in which people act and communicate with each other in real-time via digital representatives (“avatars”). Though the initial hype created by the metaverse has moved on to generative AI after the release of ChatGPT in October 2022, the social virtual reality of such metaverse remains a growing field, and several technology companies such as Meta, Microsoft, and Apple feed large amounts of money in its development, while using different foci and verbiage.

The metaverse and its associated virtual worlds will provide a three-dimensional digital space in which numerous EWOM exchanges will take place, and I assume their importance for decisions inside the metaverse, but also outside of it to rise enormously. How can those conversations, which will often be verbal, not textual, be measured? Raoul Kübler and I have just made a first empirical step into understanding and measuring such “metaverse WOM”, but way more research is needed here. How can we measure what is said from one avatar to another avatar in virtual worlds?

For leading retail platforms such as Amazon, EWOM has not only been a major driver of success in the past, but it is also an important tool for securing future success. The reason is that EWOM follows a network-effect logic, where size leads to even bigger size, whereas most other reasons to visit a retailer follow a more traditional economy-of-scales logic.

But the value of EWOM on retail platforms has been massively compromised by fake reviews recently, which in some product categories even dominate authentic reviews, making the EWOM databases essentially worthless for consumers. It is my impression that Amazon and others need to invest substantially more resources to fight the spread of such fake EWOM, as they are eventually devaluing one of their main competitive assets. Understanding the true value of EWOM for a platform and finding a way to combat fake EWOM are among the most exciting research topics. A growing stream of marketing research can help them on their way; the review on the issue by Moon and colleagues provides a good starting point.

Finally, for consumers the greatest challenge today with regard to EWOM is to find a way to cut through the growing clutter of misinformation. As the New York Times describes it:

“Fake reviews are so pervasive that nearly every online shopper has most likely encountered one. Amazon said it had blocked more than 200 million suspected fake reviews last year, and Google said it had removed 115 million rule-breaking reviews from Maps in 2022 — an increase of 20 percent from the previous year.”

Developing tools and approaches that help them to identify fake EWOM would be extremely valuable, not only from a scholarly perspective, but also from a societal one. The mere economic importance of EWOM now is the greatest challenge that governments, consumer organizations and others who want to help them have to address, if they want to prevent the shift in power from firms to consumers that the internet has brought to be reversed. I assume that in addition to legal tools, the potential of artificial intelligence are far from being exhausted in this area, but like most technological innovation, they will work in both directions — for those who try to help consumers, but also for those who want to cheat them.

Overall, we have seen EWOM to rise from rags to riches, from an outsider and underdog in its early stages to a real business champion — the Rocky Balboa of the digital world, if you want. But those who know the saga are aware that champions can become vulnerable, not because of the inherent strength of their opponents, but because of losing the “eye of the tiger”. This is what we are witnessing with EWOM these days, with the concept being under threat to lose its value for consumers and society, essentially as a result of its own huge success. Those who want EWOM to escape this threat and bounce back in dedicated Rocky style will have to turn not only to our governments, but also to retailers such as Amazon and ask them to do more.

In any way, we, as scholars, will watch and continue to probe how the EWOM version of the saga continues.

[1] See for example, Hu & Kim (2018) in the International Journal of Hospitality Management.

[2] As of January 4, 2024, analyzed with Harzing’s “Publish or Perish” app based on Google Scholar data, using the term “word of mouth” as search criteria; Google Scholar itself listed 10,032 cites of the article by other academics at that date. See also Bhaiswar et al.’s (2021) bibliometric analysis of the EWOM literature.

Let me express my gratitude to my fellow seminal EWOM explorers (and co-musketeers) Johnny Walsh, Kevin Gwinner, and Dwayne Gremler for going on this very particular journey with me. Our lasting friendships are highly valued.