Cheers in UK: How Visible Are Spanish Sparkling Wines on Google.co.uk?

In this chapter, we apply qualitative and quantitative analysis techniques with the aim of measuring the visibility of websites in searches for Spanish cava or sparkling wine made using the UK version of Google. Using the Google Ads Keyword Planner, we retrieved a group of searches in the semantic field of ‘ sparkling wines ’ and monitored the results. Extraction and also cybermetric analysis of Search Engine Result Pages (SERPs) using SEO techniques were used to calculate the visibility of Spanish cava brands via their own websites and e-commerce websites, whereupon we were able to establish rankings of media, social networks, wine sites and e-commerce websites as well as recommendations for content optimization.


Introduction
In recent decades, the wine sector has undergone major structural changes due to globalisation and increased international competition [1]. It is a highly fragmented sector where exports from the traditional producer countries (France, Italy and Spain) are being challenged by new producers, such as Australia, California, Chile, Argentina and South Africa, the latter dubbed 'New World Producers'.
Consumers are also changing globally. This is not only a geographic shift in terms of where exports are being directed, with the emergence of new markets such as China, but also a shift in terms of consumer tastes, whereby they are now opting for better quality wines and the satisfaction of more complex needs [2], something legitimate and linked to the cultural attributes associated with the product itself.
All these changes have had an impact on production, with numerous wineries, territories and countries battling to position their wines worldwide. Hence Denominations of Origin (DOs) are created, which help to certify the particular quality of wine products, whose production is limited to certain geographical areas, which in turn protects them from competition and fosters their promotion, status and international recognition. The most globally recognised DOs of sparkling wine are French Champagne, Spanish Catalan Cava and Italian Prosecco.
Although bottled still wine was the highest selling product worldwide in 2017, with a market share of 71% in value and 54% in volume, sparkling wines are gaining ground, with exports rising globally by 11.2% in volume and 8.9% in value; although they only represent 8% of the total exported volume, in terms of value that figure rises to 19% [3].
The main exporters of sparkling wine worldwide are Italy, with a market share of 41%, followed by France, with 24% and Spain with 22%. However, the average value of an exported bottle of Italian sparkling wine is €2.80, while that of Spanish sparkling wine is €1.84, and that of French is as high as €12.60, due to the high average price of champagne, €18.20 a bottle [4].
Champagne is a sparkling wine produced using the champenoise method whereby the second fermentation occurs inside each bottle, thus obtaining a superior quality product. Such high quality, status and global reputation make it the indisputable market leader. In 2017, global consumption of champagne amounted to 40% of the value and 13% of the volume of all sparkling wine, with France and the UK being the main consumer markets for this precious liquid [4].
Catalan cava is an alternative to champagne. It is made by the same production process as champagne, but the price is lower, so it is highly appreciated on the market for offering good value for money. Because of this, exports have grown exponentially since the 1990s to reach 162 million bottles in 2017, with sales led by the Freixenet and Codorníu companies.
Likewise, in recent years, Italian Prosecco has become one of the leading sparkling wines on the international market, regarded as a fashionable but cheaper alternative to French champagne. It costs less than champagne and cava because it is produced using the charmat method by which the second fermentation occurs in stainless steel tanks, which brings down production costs in comparison with the champenoise method. In 2016, Prosecco exports to the UK exceeded those of champagne in terms of value, and forecasts suggest it will become the fastest-rising sparkling wine in terms of worldwide sales [4,5].
As for the main countries that import sparkling wine, the UK tops the ranking in terms of value, playing its part in the upward trend in sparkling wine consumption, with a 76% increase in sales from 2012 to 2017 [4]. In 2015, 80.4 million litres of Italian sparkling wine were sold in the UK, followed by the 28 million litres of French sparkling wines and 20.2 million litres of Spanish sparkling wines [6,7].
The Spanish wine sector, through the internationalisation of its wineries and designation of origin (DO) wines, is highly prestigious and internationally renowned. It is a key sector of the Spanish economy, culture and gastronomy, and has now reached full maturity [8]. However, the economic crisis that began in 2007 led to a decrease in domestic sales and stock accumulation, leading many companies, especially SMEs, to begin or intensify internationalisation processes (ICEX, 2013).
In this chapter, we shall be analysing the visibility of websites when searching for Spanish cava or sparkling wine using the UK version of the Google search engine.
In consideration of technological progress and the increase in online sales, a recent report by the RAW agency on 'The digital behaviour of Spanish wineries' [16] states that only two out of every 10 Spanish wineries get more than 10,000 visits to their website per month and in most cases these are limited to visits from inside Spain. In turn, a large proportion of Spanish wineries' foreign sales are made through wholesalers, since most of their websites are not addressed at the international digital market or at direct sales of their products. All of this suggests that there is a huge opportunity for them to increase their sales through a greater presence in the digital market, especially in the UK market that tops the rankings for the online sale of alcoholic beverages.
Estimates of international traffic generated with the SemRush digital marketing instrument show that only seven of the 10 highest selling Spanish wineries receive traffic from the UK ( Table 1).
Although the instrument does not define the types of traffic (whether it is direct, via social networks, via links in newsletters or via search engines), the purpose of our study is to measure a website's visibility based on of the contents that appear on Search Engine Result Pages (SERPs).

Research goals and questions
The main objective of this study is to analyse the visibility of the Spanish sparkling wine brands on the UK version of the Google search engine, that is, to identify the websites that appear in the results of searches for 'cava' and thereby determine how well positioned Spanish cava brands are.
Our specific goals are: 1. Mapping of all websites visible on Google.co.uk when making searches related to Spanish cava or sparkling wines, in order to identify websites that, to a greater or lesser extent, produce content that appears on SERPs in response to searches related to 'cava'.
2. Analyse the Google visibility of the websites of Spanish sparkling wine brands. By mapping, we are able to identify sparkling wine websites and classify them by nationality, in order to accurately measure the visibility of websites belonging to Spanish brands and/or with a Spanish IP address.
3. Analyse the Google.co.uk SERP visibility of press, social network and sector website results in order to measure their weight and importance in the organic results, thus recognising the websites that are best rewarded by Google in these categories. 4. Analyse the visibility of retailers that appear in Google results and which offer Spanish sparkling wine brands, from specialist stores to major supermarket chains.
Bearing these objectives in mind, the research questions that this study will be looking to answer are the following: RQ1. RQ2. What is the relative weight of the websites of Spanish sparkling wine brands in Google.co.uk results?
RQ3. What media, social networks and sector websites are shown by Google when searching for Spanish sparkling wines from the UK? What is their relative weight?
RQ4. Which are the main e-commerce websites in terms of their visibility?

Conceptual framework
It is well known that the vast majority of Europeans routinely use search engines to access the information available on the web, and that the links appearing on SERPs are their main point of access to the web immediately after performing a search using their search engine of choice. In the UK in 2018, Google had a market share of 89.1% of all searches made on search engines, thereby acting as gatekeeper [9] in terms of access to information, meaning that its search engine results pages are the best showcase for companies seeking to promote their products or services.
Search engine results pages (hereinafter SERPs) are one of the central objects of study in Search Engine Optimization (SEO), because they reveal what Google 'considers' to be the best possible response to the searches made by users, and if a website is viewed by Google to be one of the best answers to a query, that page will be ranked high on SERPs and will therefore benefit from the enormous flow of traffic that stems from being in the top positions.
At first, Google's SERP offered 10 organic results, that is, 10 links to web content that had previously been located by Google crawlers and that the search engine ordered by how relevant they were to its algorithm's criteria. However, since the incorporation of the Google Universal and Blended Results algorithm [10], its SERPs have undergone constant change and currently consist of a variety of elements that appear along with the 10 original organic results. These elements include images, videos, news, local results, social network content and special results generated by the search engine itself based on web content that it has been discovered during its page crawling and indexing processes. The most prominent of these elements include the Knowledge Graph [11], 'One Box' responses and, since 2018, the 'Related Searches' that appear in more than 80% of the searches we make.
However, despite the many elements that the search engine displays on its results page, as many as 84% of clicks are concentrated in the first four results [12], which makes the highest part of the SERP an extremely competitive space, not only for the websites that are looking to gain traffic by appearing in the top results, but also for the search engine itself, which fills up much of this space with its own elements.
With the aim of cataloguing and measuring the visibility of contents appearing on SERPs when performing searches related to Spanish cava and sparkling wine, the following study analyses SERP composition using the Infosphere model proposed by Peretti [13] and Bennett [14]. This media ecology applied as an open system of categorisation [15] can classify any content and measure its weight and importance on SERPs using a system composed of three well-differentiated and clearly identifiable strata, namely: • The Conventional Layer corresponds to the media and is made up of cybermedia, meaning both the conventional press, and any other newspaper or magazine that is published on the web.
• The Middle Layer consists of millions of websites and includes leading domains on any subject.
• The Micro Layer includes social networks and the blogosphere.
Each of these layers contains different sectors and conglomerates, groups of domains that all share a characteristic sector or theme. The system includes a database of websites thanks to which it is possible to identify and measure the proportion of websites appearing on SERPs individually or together with other sites in the same category. In addition, as it is an open system, the database can be expanded with new thematic conglomerates depending on the nature of the research, as we have done with our sample.

Methodology
This research is a longitudinal study over a period of time that analyses the composition of the SERPs shown by Google when we search for a specific topic, in our case related to the Spanish sparkling wine. The study combines quantitative and qualitative methodologies to analyse a series of samples of potentially visible SERPs in searches conducted from the UK. The qualitative analysis allows us to recognise different types of results shown by Google in its SERPs, as well as the different types of websites depending on their characteristics and themes. On the other hand, a quantitative analysis is applied to the categorisation derived from the qualitative analysis, thanks to which we can measure the weight of any of the sample categories.
Thus, the present study analyses a sample of Google.co.uk results found between May and November 2018 using a set of keywords related to 'cava'. The retrieval of these terms together with their frequency per month is the first step of the methodology, a list of keywords used in the UK to find information about sparkling wines or to buy them.
Using the Google Ads 'Keyword Planner' tool, previously known as Google Adwords, we were able to extract lists of keywords related to any term along with its monthly frequency in the last 4 years. The interface can be used to select any country in the world and extract results that are broadly or closely related to the initial search term.
In order to find out the search frequency of the different terms associated with this search niche, we checked four 'seed terms' in the planner: on the one hand, the term 'cava' together with its biggest French and Italian competitors, 'champagne' and 'prosecco', and on the other hand, the generic term 'sparkling wine'.
These four initial terms, after several iterations, produced a total of 2280 queries that had gathered an average of 481,640 searches per month during the previous 4 years.
There were 752 retrieved queries including the term 'prosecco', which accumulated an average of 215,000 searches a month, followed by those that include the term 'champagne', of which there were 691, but which accumulated a higher average of around 224,000 searches a month. At a considerable distance behind, we were able to identify 289 searches that included the term 'cava' and which accumulated around 22,000 searches per month, a lower figure than the 682 queries in which the term 'sparkling' appeared, which accumulated about 24,000 searches per month. Also, and to a lesser extent, 42 related terms were identified that include the term 'fizzy' and accumulated about 1500 searches a month (see Table 2).
In the aforesaid large groups, we can find searches that include more than one of the national terms that we use to describe sparkling wine. Searches that include 'champagne' and 'prosecco' have an average monthly frequency of 2610 searches, followed by those featuring the term 'champagne' together with 'cava' with 700 searches a month and 'cava' and 'Prosecco' with 640.
Once the keywords had been retrieved, we selected a subset of 24 representative terms, including the term 'cava' along with various search modifiers, which we proceeded to use as the main terms for extracting Google UK search engine results, both organic and those paid for using the Google Ads service. Table 3 shows the keywords used along with their monthly and accumulated average frequency for the years 2016-2018.
As mentioned before, we start from a database of more than 60,000 sites, which include media websites from different countries, the most popular online social networks, as well as relevant sites and blogs on a variety of topics. We can use this database to take any sample of results and identify and measure those that belong to the Conventional Layer or the Micro Layer. Depending on the subject of the seed terms, each new capture of results presented new websites that had not been identified previously and that needed to be categorised.
The new capture of results with a set of terms related to cava allowed us to locate unknown websites that are categorised via a semi-automatic process that recognises the type and topic from the domain name, elements of the Uniform Resource Locator (hereinafter, URLs) and the metadata that we extract afterwards. These new websites mostly belong to the Middle Layer, especially when the thematic niche had not been mapped before.
Once the keywords had been selected, we applied Search Engine Optimization (hereinafter, SEO) tools to extract the Google.co.uk results for each term. The instrument used in this phase was Advanced Web Ranking Cloud, which was programmed to extract both the Top5 SERPs from the Mobile and Universal (desktop) versions, and results for paid Google Ads, which include advertisers who have bid for each of the terms used.
The data capture period lasted for 29 weeks from May to November 2018 (see Table 4), resulting in a sample of 72,302 organic results taken from Google.co.uk Mobile (36,316 URLs) and Google.co.uk Universal (35,986 URLs), and 2193 results from Google.co.uk Pay per Click (Google Ads). After retrieving the visible URLs from the organic results and advertisements, the parameters of the URLs were deleted and the domain names isolated, enabling us to identify a total of 757 different websites and 59 advertisers.
To measure and analyse the size of the different sectors and conglomerates, we have a database of previously categorised sites, a pre-established group of homogeneous conglomerates common to samples from different niches (press, social networks, government websites, etc.). However, during this data capture process, new sites were discovered and to categorise them, new conglomerates were created that fit specific types of websites that appeared in our sample, such as international sites of sparkling wine brands, liquor stores, websites and blogs on the wine world and a series of lesser sites that showed up in the sample. In this case, a database of national and international websites linked to the wine world was developed that includes more than 4000 Spanish wineries and thousands of websites and blogs.
Then all the results with visibility were categorised by layer, sector and conglomerate, followed by analysis of the complete dataset according to type of site, which of these categories they belong to and their position.
All the visibility calculations were performed on Excel spreadsheets, and dynamic tables were used to analyse the subsets of data.
To summarise, the following process was observed in order to answer the research questions: 1. Extraction of a list of searches made from the UK related to sparkling wine and cava using the keyword research features of the Adwords and SEMrush instrument. 2. Selection of a set of 24 representative searches.
3. Scraping of the first five pages of results of each search on Google.co.uk using the Advanced Web Ranking Cloud instrument.
4. From the resulting package of SERPs, the URLs were extracted and the domain names to which they belong were isolated.
5. The domains were categorised by type and grouped into homogeneous conglomerates.
6. The composition of the results of the sample was analysed.

Results
The top five SERPs for each term were retrieved using scraping processes, extracting a sample of 74,495 URLs from Google.co.uk Pay per Click, Google.co.uk Mobile and Google.co.uk Universal. Categorisation of the domains of all the URLs in the sample enabled us to generate a database of websites that could provide an answer to the four research questions about the visibility of websites in the Top4.
The first research question asked what kind of information Google.co.uk shows when British users search for Spanish sparkling wines. Table 5 shows the distribution of results by device used and the type of content that appeared in the Top5 SERPs. Most of the retrieved contents are organic results, 77.23% of the total, followed by images with 7.73%, Accelerated Mobile Pages (AMP) only detected on Mobile with 7.15%, and Related Questions with 5.98%.
The weight of the remaining content is considerably lower, as shown in Table 5. Videos, mostly from YouTube, account for 1.08% of the total, and 'One Box' only amounted to 0.75%.
The aforesaid proportions change substantially when we focus on the results that appear in the visible part of the SERP, that is, the top four results. Table 6 shows the same distribution as in the previous table but only for the Top4.
When analysing this sample we note that the organic results represent only 32.79% of the results (28.74% on mobile and 36.85% on desktop) and the remarkably high visibility achieved by the Related Questions section in the top part of the SERP with an average of 43.86%, while that figure rises to 48.96% in the desktop version. The sample hence evidences the visual preponderance of these questions generated by Google over the other contents, even above the organic results. Next, we applied the different Infosphere levels to the sample of results for the Top4, which allowed us to group the different websites according to which of the three main layers they belong to, and at a lower level by grouping them into sectors and conglomerates (see Table 7).
On the first level of analysis, we distinguish the three layers into which the Infosphere model is divided. The results from the Conventional Layer make up 56.48% of the total (3.145 URLs), followed by the Middle Layer with 34.72% (1.933 URLs) and the Micro Layer with an average of 8.80% (490 URLs). However, we note that the proportions of the first two layers vary considerably in terms of Mobile and Desktop results, with a difference of 10 points more for media content in the Desktop version.
On the next level of analysis, we find a subcategory called Sector that encompasses sites that share common characteristics either in terms of sector or topic. In the Conventional Layer, we find two very distinct sectors, results coming from the press on the one hand, and results generated by Google on the other. On this level, we observe that the results generated by the search engine accumulate 43.89% compared to those by the press, which only amount to 12.59%. The most prominent sectors in the Middle Layer include the group of sites dealing with e-commerce (14.03%), followed by websites on various topics (13.06%) and sites about the wine sector (6.72%).
Finally, we calculated the proportions on a conglomerate level, where 43.89% of the search engine's results are 'Related Questions', followed by the British Press conglomerate with 12.07%. In third place, we find the Middle Layer conglomerates that group food and drink websites with 10.65%, followed in fourth and fifth place by the conglomerates that group Supermarkets and Liquor Stores with 8.21 and 3.95%, respectively.
Focusing now on the second research question that enquires about the weight of the websites of Spanish sparkling wine brands in Google.co.uk results, we observe that the visibility of the Spanish brands that appear in UK results of is practically non-existent (see Table 8). The Freixenet brand stands out as the undisputed leader in terms of visibility both for mobile and desktop searches with 1169 ranked URLs, 1.5% of the total sample. The brand has the advantage of owning as many as four different domains, freixenetusa.com, freixenet.com, freixenet.es and f reixenet.co.uk, the first of which features in 19 of the 24 searches. Its biggest  Table 7.
Proportion of layers, sectors and conglomerates in the Top4.
beneath the top of the SERPs and whose visibility and click probability are almost 0%. It should also be noted that the only specific UK domain, freixenet.co.uk, appears in average positions of 43 and 46 out of 50, with virtually null visibility. These average URL rankings of the 10 Spanish brands in 25-46th place out of 50 confirm the information published by the RAW agency on the digital behaviour of Spanish wineries, claiming that they have not been launched onto the digital market. This is why we advise these companies to work to get their websites into the Top10 of Google search engines, through SEO optimization or payment for keywords to appear in SEM, and thus achieve greater international visibility for their brands and products, focusing first on the international markets where the company has the highest sales and, secondly, on developing markets where there are good sales prospects.
The third research question asked what media, social networks and wine sites are retrieved by Google when searching for Spanish sparkling wines from the UK and what their weight is.

Media visibility
After performing an analysis of the media that appear in the Top4 Google.co.uk positions (see Table 9), we note that 43.86% correspond to questions that the search engine suggests to users who are looking for information about Spanish cava or sparkling wine.
In a prominent second place, we found the independent.co.uk newspaper, which covers 7.17% of searches made, followed by the telegraph.co.uk newspaper with 2.3% and theguardian.com with 1.90% of results when combining results for its website and the AMP mobile version.
From this analysis, we observed that these Spanish cava companies' marketing plans do not include public relations or advertising actions with these media, since their visibility is very limited.

Visibility of social networks
Regarding the visibility of social networks in searches related to cava or sparkling wine (see Table 10), we should note that only en.wikipedia.org and youtube. com hold any significant weight in the Top4 with 5.85 and 2.86%, respectively. Other social networks have only marginal visibility, since twitter.com and eyeonspa in.com each has two ranked items of content, and m.facebook.com only has one. These results suggest that Spanish cava companies do not prioritise communication of their products on the web pages shown in Table 10 and from which they could obtain much better visibility among British users.

Visibility of wine sector websites
As shown in Table 11, the leading wine website in terms of visibility is winefolly. com which accumulates 2.71% of the total results in the Top4, claiming top spot on Mobile (4.53%) and coming third in Universal (0.90%). The next sites in the global ranking are vinepair.com and winepleasures.com with 2.19 and 0.97% respectively, although the former site has a greater number of ranked content in the Universal version. The fourth and final question asks what the main e-commerce websites are in terms of their visibility.
As we mentioned earlier, by mapping websites on the sample of results, we can measure the visibility of any website whose contents are visible on Google SERPs. The database resulting from this mapping procedure includes a series of e-commerce sites that we classified into three different conglomerates based on the products offered. So, we have a first conglomerate where all the results from Ama zon.co.uk were grouped, a specific conglomerate of supermarkets with visibility in the Top4, and a third for e-commerce websites specialising in wines and spirits. Table 12 shows all the e-commerce websites that appeared in the sample grouped into the three aforementioned conglomerates.
The analysis also shows that the e-commerce domain that appears most frequently and has the best SEO rankings is tesco.com, as it appears 210 times (3.77%); in second place is groceries.morrison.com with 144 URLs, followed by the laithwa ites.co.uk wine store with 75 URLs and groceries.asda.com with 57, while Amazon. co.uk appears seventh with 48 URLs ranked in the Top4. It should also be noted that the m.tesco.com link appears 48 times and sends users directly to the supermarket's wine section.

Conclusions
In this chapter, we have quantified a representative set of queries related to sparkling wine used by British users when searching for information on Google. From this set of queries, we have selected the ones that are directly linked to cava and extracted search engine results for a 29-week period so as to generate a database of all visible websites on Google.co.uk when searching for Spanish sparkling wines, among which we have identified the websites of Spanish brands.
Using the Google Ads Keyword Planner instrument, we retrieved the monthly frequency of a set of queries related to sparkling wine that include the generic term 'sparkling', the word 'cava', or the names of its direct competitors, namely 'champagne' and 'prosecco'.
From analysing this frequency data, we could confirm the rising tendency of the term 'prosecco', which is the most in-demand sparkling wine on international markets and the one that accumulates the highest number of queries in the UK; we also confirm the hegemony of 'champagne', which accumulates the highest number of searches a month, with searches for 'cava' and 'sparkling wine' in third and fourth place.
After this initial analysis, a set of 24 keywords related to Spanish sparkling wine were selected and SEO techniques were applied to extract the URLs that are shown when British users search using these keywords. The result was a sample of URLs and websites with which we could map all the websites that are visible on Google.c o.uk for searches on subjects related to Spanish cava or sparkling wines.
Based on this mapping of results from Google.co.uk, we performed a cybermetric analysis of the web domains to work out the visibility of any website associated with the Spanish sparkling wine niche, and in particular to quantify the visibility of Spanish brands on SERPs. This is how we were able to measure the visibility of the websites of Spanish sparkling wine brands, noting that no cava brand is ranked in the Top4, and that the site with the highest visibility, freixenetusa.com, reaches an average ranking of 25, a position with a visibility and click probability of almost 0%. On average, the 10 Spanish brands detected in the sample are in around 25th and 46th place, which confirms the data published by the RAW agency that claims that these brands have not been launched on the digital market.
Likewise, all the web contents that appear in the results have been identified, especially those belonging to the media, social networks, wine websites and ecommerce sites. These data can be used as the basis to suggest content strategies that will gain visibility in the search niche, and offer guidance on where Spanish brands should focus in order to increase their visibility among British users and overcome barriers for the sale of wine.
Finally, we should stress that the mapping of websites in the sector shows what contents are chosen by the Google algorithm from among thousands of possible candidate sites. This selection showcases the best contents that any winery should take into account when creating their digital identity.
As a final conclusion, we could comment that if Google's Consumer Barometer mentions that 'By understanding what consumers look for, businesses are able to ensure the right information is available on their websites', by understanding what results appear on Google, we can make the contents available on cava websites visible from any search engine.

Implications for managers
The managerial applications derived from this study are aimed at improving the marketing strategies of Spanish wineries. The results of the research can help to optimise a company's digital identity and lead to a variety of actions aimed at improving not only the visibility of its website but also sales of products.
The following are four particular actions derived from this study whose real application will help to boost the ranking of cava brands and products.
Our first recommended action is to improve SEO strategies, that is, optimise websites to position their contents at the top of SERPs and thus achieve greater international visibility for brands and products. Brands should first focus on the international markets where the company has the highest sales and secondly, on developing markets where there are good sales prospects.
The second action is aimed at improving knowledge about consumers through search engine analysis. On the one hand, quantitative analysis of the frequency of Google queries related to cava, champagne and prosecco can be used to measure the level of relative and absolute interest in these three products that compete not only in the British market but also on a global scale. On the other hand, qualitative analysis of long-tail queries offers specific details about the perception of cava, prosecco and champagne among the population of any country.
The third action is directly related to improved website visibility, as knowledge of the ranking and level of optimisation of a website can reveal whether the content marketing plan is working in a certain market. That is why this third action is based on finding out what types of content are preferred by Google, which will help us to develop optimised content strategies based on the data extracted from the sample.
In the case of cava, it has been detected that independent.co.uk, telegraph.co.uk and theguardian.com are the press websites that cava brands should focus their public relations, advertising and press releases on, since they are the sites that can capture the most organic traffic for searches related to cava. Likewise, the social networks that have been identified as having the greatest visibility in British results, and for which we advise companies to develop good content, are Wikipedia, YouTube, Twitter and Facebook; to feature on the wine websites with the highest visibility in the UK, they should target efforts on winefolly.com, vinepair.com and winepleasures.com, and negotiate linear marketing and promotional deals with the leading e-commerce sites, particularly tesco.com, groceries.morrison.com, laithwa ites.co.uk, groceries.asda.com and totalwine.com, as these will not only take charge of selling the products, but also advertise and promote the brands and products that they work with on their own web pages. Knowledge of these media, blogs, websites and e-commerce companies will help to plan content, advertising and brandedcontent strategies to achieve better brand and product visibility in the eyes of British netizens.
In conclusion, we are able to make recommendations to improve the digital identity of cava brands aiming to gain visibility in a specific market.

Limitations
This study is not free of limitations. First, the selection of a single national search engine tells us about the visibility of Spanish sparkling wine sites in that specific country, but the results cannot be extrapolated to markets other than the UK. Secondly, although the AWR instrument can be used to select IPs from a specific country, the personalization of Google results based on the user's location can cause variations in the rankings and proportions of results for the same search. It would therefore be advisable to perform the data extractions from different locations in the UK and compare the results. Thirdly, constant modifications to Google's ranking algorithms can cause considerable variations in the results, so this study is only valid in the short or medium term. Therefore, the recommendations derived from these analyses need to be adjusted to further monitoring of the Google rankings over time in order to cater for evolutions in terms of results.

Future lines of research
Future research derived from this study includes increasing the sample size by expanding the list of analysed terms and/or analysing a higher number of locations.
This study was conducted on the basis of a limited group of searches, that is, a set of keywords limited to 24 generic terms related to cava and Spanish sparkling wine. However we could expand the list of terms in different ways by adding new combinations of keywords belonging to the same semantic field.
For example, we could add new generic searches related to the different types of cava and their variants ('cava' + 'type of grape' / 'cava' + DO / 'cava' + 'brand' / 'cava' + 'name of supermarket', 'cava' + 'search transaction modifiers') and measure the exact, real-time visibility of Catalan sparkling wine brands in Google results viewed from the UK or from any other country. Another option would be to use a set of keywords that includes terms related to wine tourism in Spain, which would allow us to measure the visibility, from any country in the world, of websites selling trips to Spanish vineyards.
The exercise could also be replicated with new sets of keywords using generic searches related to 'champagne' and 'prosecco', in order to measure the visibility of Italian and French wineries in the UK or on any other search engine.
As well as keywords, we could also increase the number of locations in the UK in order to measure deviations from this data capture, or choose any other Spanish wine-buying market by selecting different Google search engines designed for specific countries or cities.
Also, and in a similar way to what the SEMrush instrument does, at present we can only estimate the traffic from Google.co.uk to Spanish wineries in terms of the amount of content that appears on SERPs and their position. In future research, it will be essential to contrast visibility on Google.co.uk with the real traffic data of Spanish sparkling wine sites. Doing so would necessarily require traffic data on Spanish wineries obtained by accessing the Google Analytics platform of the different websites.
In the long term, with a large database both of terms and sector websites, an observatory of the wine search niche focused on measuring Spanish websites in potential importer countries could be created.