Repurposing Google Ads, The Frames of Influencer Marketing

Exploring how influencer marketing is framed by Google ads as a form of gig work.

Team Members

Sheng Wong, Noyan Er, Arturo Arrigada

Project slides
https://docs.google.com/presentation/d/1F3JTozz3F53dWL1_io5Famwxi84_uMs5z8JjCUn1O3Q/edit?usp=sharing

Summary of Key Findings

Companies or brands are now starting to use influencers as a marketing tool via social media, which is becoming an increasingly effective instrument that has seen its market size grow dramatically (Craig, 2019). Nevertheless, how do influencers find this type of work through Google? Our research explores how Google ads frame the topic of influencer work as a type of gig work, examining its queries to assess the association of actors as well as queries prompting questions of concern.

Our findings show that news outlets held a majority of URL results on search engine pages when concerned keywords were utilised. Out of the concerned queries, 35 were associated with news outlets that expressed concerns about mental health. Moreover, agencies and intermediaries dominated queries that were utilising keywords in regards to marketing. 95 out of the 101 URLs that were investigated conveyed that 27 of the URLs were associated with intermediaries and blogs also featured heavily in this category.

Furthermore, the findings suggest that intermediaries and agencies are paying to appear in marketing queries as well as concerned ones. It delineates how heavily reliant actors are in positioning themselves through paid and organic queries of their interest.

1. Introduction

How does an influencer search for work on Google? With the right keywords, companies and agencies related to influencer marketing will pop up as Google advertisements in the search results. But how are these keywords framed by Google Ads in regards to work for influencers?

Google Ads is an online advertising program that was launched in 2000, where the creation of online advertisements for products and services are reflected within the search results. Advertisers use this tool to compete for search terms and adverts can be applied locally, nationally, or internationally. A controversial study that looked into how Google Ads repurposed the information related to climate change found that 78 of the climate-related terms were placed by firms with an interest in fossil fuels. It found that ExxonMobil, Shell, Aramco, Mckinsey and Goldman Sachs, all well-known polluters and big investors in fossil fuels were among the top 20 advertisers in search queries. (“Fossil fuel firms among biggest spenders on Google ads that look like search results”). If Google Ads is able to repurpose information in this manner, then it raises the question to what extent can Google ads be used as a legitimate tool to represent information?

The term ‘Influencer’ had only become prominent with the rise of social media which has become synonymous with content creators who gather large numbers of followers, deeming them ‘insta-famous’ (Instagram-famous) (for more examples see Cunningham & Craig, 2019; Duffy, 2017). This growing trend of using influencers to market brands has become an effective tool because it fosters a parasocial interaction (PSI) with its followers. One study that evaluated this subject matter found that PSI was positively determined by an attitude of homophily and social attractiveness when it came to influencing followers purchasing decisions. (Sokolova and Kefi 2020).

However, why is influencer work considered to be gig work? Watson et al (2021) describe the following three themes in order to define the characteristics of a gig-worker: Firstly, work that is enabled by a technology “that facilitates the matching of supply and demand for services” (Watson et al 332), For example, promoting via Instagram or Tiktok. Secondly, work is arranged not for the long term, nor do these workers receive a salary or a wage, for example, influencers are typically freelancers. Thirdly, these workers are defined solely through their tax status or legal classification, for example, since influencers side more with freelancers in the legal aspect, they would receive a different tax form. Influencer labour has been described also as an activity based on the promise to monetize individuals’ activities of content creation, even though not all of them are capable of “making a living” through this activity (Duffy, 2017).

An industrial report conveyed how influencer marketing had increased in 2021 despite the threat of COVID. The market size was set to jump to $13.8 billion compared with its $9.7 billion market size in 2020. Furthermore, new channels of marketing such as Tiktok saw influencer usage increase from 16,394 in 2020 to 35,528 in 2021, emphasising its importance as a marketing tool. (“The State of Influencer Marketing 2021: Benchmark Report”).

2. Research Questions

There are three research questions within the focus of this study, aims to elaborate the interactions between potential influencers, content creators and the different actors they can possibly encounter while following their means to monetise their social media presence.

• How “influencer marketing” is framed on Google ads and landing pages? Specifically, for two types of actors: influencer marketing agencies (e.g. companies that intermediate between brands and influencers) and potential influencers (e.g. people who want to work as an influencer).

• What kind of concepts (e.g. keywords) describe these activities as a new form of gig work?

• How economic and social value -from the perspective of different actors- is defined and/or contested?

3. Methodology

Our analysis looks to evaluate what landing pages and keywords were utilised in order to draw audiences in the field of influencer marketing using the following tools: Ahrefs, and Answer the public. The initial datasets that were collected mainly targeted queries by adding well known social media networks such as Instagram or Tiktok. Answer the public was also used in order to explore what potential queries would be effective in our keyword search.

The methodology was designed to start off with the construction of a set of keywords after reviewing relevant literature. The answerthepublic tool was used in order to gather other relevant keywords in the forms of related questions around the initial set of keywords. The final data collected from the answerthepublic tool was used in the ahrefs tool where the keyword explorer function was used to list results that were positioned in search engine result pages yielded by these queries. The site explorer function of ahrefs then enabled us to find out for which other keywords these domains were shown for in the first page of the search engine results. This re-iterated large set of keywords was curated in order to select the non-repetitive keywords which would fall in the scope and aim of this study. Finally, we used the site explorer function again to list the URLs that were placed within the first page of the search engine results in order to identify singular actors.

Our search focused only on the United States because we believe that the influencer culture would be more prominent compared to other countries, allowing us to yield more interesting results in a diverse population where the consumer culture is increasingly prompted by social media.

3.1 Categorisation of keywords

We decided to focus on collecting paid keywords and organic queries from the United States using Ahrefs. These keywords would be categorised in two specific fields: Marketing & Concerned

Marketing

The keywords that reflect queries that look to monetise an individual's presence on social media e.g influencer marketing company or easy sponsorships.

Concerned

The keywords that reflect concerns revolving around the issue of influencer marketing e.g social media burnout or influencers mental health.

3.2 Identifying actor types and visualising the data

To determine what actor types were present, we assessed the landing pages of all the search queries we gathered using textual analysis and categorised what types of actors, for example, agencies, companies, blogs, and etc. In total, we identified seven types of categories of actors and categories each actor accordingly:

Agency: Financial entity with an emphasis on building a portfolio of influencers to be utilised in the marketing industry.

Intermediary: Financial entity providing technological solutions for an aggregate of influencers/brands in terms of campaign building, analytics and monitoring.

Media Outlet: Content aggregator which operates beyond the scope of providing contemporary news. Usually contains links to several content platforms.

News Outlet: General news source which focuses on many kinds of contemporary issues.

Blog: A content outlet that utilises content created or curated by a single figure who is a type influencer themselves.

Health Service: Government or private entity that provides health services.

Platform: Multi-faceted entity whose activities go beyond influencer/content creator related production. Widespread social media and content platforms such as Instagram and YouTube fall under this category.

For the visualisation purposes of the finalised dataset, we used the tool RAWGraphs (https://rawgraphs.io/) that provides us with different options for representations with regards to hierarchies and proportions of the final data set. From the different options the ‘’Sunburst diagram’’ graph type was selected as we were interested in a comparative representation of different actors with regards to how heavily they accumulate the results with regards to different types of queries.

4. Findings

Our findings consisted of different types of queries, the URLs that were included in the search engine result pages were included in these, how these were correlating and which types of actors were prominent in search engine result pages yielded by different types of queries.

While scaling the concerned keywords by utilising the online topical research tool answerthepublic, (https://answerthepublic.com/) it has come to our attention that most of the concerned queries were addressing the mental health issues regarding social influencers, marketing professionals and social media users in general. These in turn generated search engine result pages which were heavily accumulated by news resources. Concerned keywords yielded 101 URL results in the first search engine result page. None of the concerned keywords yielded paid search engine results. Marketing keywords, on the other hand, yielded 181 URL results on the first page of which 32 were paid. These were mostly found out to be concerned about the practical issues in monetising an individual's presence on social media. Some of the marketing keywords referred to the search for an explicit affiliate in social media influencer marketing.

News outlets hold the major part of URL results in search engine result pages when concerned keywords are used as queries on Google. Out of the 101 URL results yielded by queries using concerned keywords, 35 were provided by news outlets. Among these news outlets were examples that are specialised in mental health issues and mental well being to be found. As expected, agencies and intermediaries are prominent in queries that were utilising keywords focused on marketing. 95 of the 101 URLs that were shown in search engine result pages were provided by actors categorised as agencies whereas 27 of the URLs belonging to intermediaries. Also, blogs were included more heavily in marketing queries compared to concerned queries. Of the paid results that were prominent in marketing queries, nearly all were provided by agencies and intermediaries.

Another finding that was found in the framework with the study was with regards to the types of actors that were being included in the search results that were being shown in different types of queries. As mentioned, most of the URLs that were included in search engine results yielded by concerned queries were provided by news outlets. An interesting finding though, is that URLs provided by agencies and intermediaries were also included in the search engine result pages provided by concerned queries. Out of the 101 results that were connected with concerned queries, 20 were provided by agencies. Another 13 were provided by intermediaries which indicates that all the actor types which are paying to be included in marketing queries are also providing content related to concerned searches and are being included in the related search engine result pages.

A further finding suggests that the intermediaries and agencies which are paying to be included in marketing queries are being included in the search engine result pages related to concerned queries are being shown in these organically.

The findings show us how heavily different types of search engine result pages were accumulated by different types of actors and how (paid or organic) these actors positioned themselves in these queries.

5. Discussion

The study shows that the primary issues regarding social media influencers are related to mental health. We found that the most common questions related to negative effects of social media included keywords such as ‘’social media burnout’’, ‘’influencers mental health’’, and ‘’avoiding social media’’ indicate a critical perspective on social media in relation to mental health. A study focused on the comparative nature of the usage of the social media platform Facebook indicates ‘’...that engaging in frequent social comparisons of any kind may be deleterious to one’s mental well-being’’ (Steers et. al., 2014). Together with the findings of this study, it is our conclusion that these concerns manifest themselves as queries conducted on the Google search engine. Another study indicates burnout related to heavy social media usage and comparisons between the users of social media and the actors they are confronted with on social media platforms (Petersen, 2019).

Our findings indicate that there is an activity in content creation where multiple types of actors are active in providing guidelines and suggestions on “have to tackle” these on the content creators’ or influencer’s end, suggesting that the mental issues are being acknowledged and vocalised in the form of queries.

Another implication of this study’s findings concerns the intermediary and agency actor types and their positionings around the different types of queries. We found out that these actor types are being shown in nearly all the paid search engine result positions that were yielded by marketing type queries we utilised, and that they were also prominent in organic results. It has come to our attention, however, that these types of actors were also producing content that was targeting concerned queries.

The fact that the content about influencer mental health produced by these mediating actors is shown in an organic (non-paid) way on the search engine result pages indicates that this type of content is presented in a way that would not necessarily provoke a critical response, as opposed to actors being present in critical issues with paid advertisements (McIntyre, 2022), but rather these mediators are acquiring a more underlying role in the personal life of social media influencers by providing suggestions about their personal lifestyle.

The study has some limitations in terms of providing comparative aspects of how influencer marketing is perceived in different countries and languages. This is caused by the utilisation of the ahrefs tool with regards to scaling keywords and mapping out different actors around search queries. Although the tool provides reliable results in terms of keywords and related ‘’parent topics’’ in the English language, similar data in other languages such as Spanish and Turkish were less reliable and incomprehensive. These conditions excluded the possibility of a cross-cultural comparative approach which would possibly provide data to be utilised in a more in-depth discussion.

6. Conclusions

We can conclude that the influencer industry is a result of different types of connections and entanglements between actors, technologies, meanings and commercial exchanges in the form of “communicative value chains” (Arriagada, 2019), where economic and cultural value is produced through the connections between these actors (e.g. paid advertising, content, Google, influencers, intermediaries, media outlets).

The growing interest of actors that are concerned with monetising social media presence and developing a portfolio of social influencers around issues that are prevalent among social media influencers and content creators calls for a critical approach towards these actors. Moreover, the phenomena of Google including these in the search engine results opens up theoretical possibilities in terms of platform critique.

The presence and activities of any financial entity which intends to utilise activities of social media influencers and content creators would be of value in terms of understanding and developing a critical approach towards the role of these entities. The interaction between these and social media influencers and content creators could also be analysed in terms of the power structures which are at play during these monetisation processes.

7. Data Sets

As stated earlier, the categorisation of data was differentiated between paid keywords and concerned keywords using Ahrefs. The final tab represents how actors were identified and how they were organised into groups that reveals whether they sided with the marketing or concerned queries.

The data can be seen on the following link:

https://docs.google.com/spreadsheets/d/1aqafgcCIxj8_aJ9UBBYQldnFQWAcIyef93Y42kPPpuU/edit?usp=sharing

8. References

Arriagada, A. (2019). Communicative value chains. In L. Vodanovic (ed). Lifestyle Journalism: Social Media, Consumption and Experience. Routledge, pp.

Craig, D., Cunningham, S. (2019). Social media entertainment: The new intersection of Hollywood and Silicon Valley. New York University Press.

Duffy, B. E. (2017). (Not) getting paid to do what you love. Yale University Press.

Geyser, Werner. ‘The State of Influencer Marketing 2021: Benchmark Report’. Influencer Marketing Hub, 6 January 2022. https://influencermarketinghub.com/influencer-marketing-benchmark-report-2021/.

Sokolova, Karina, and Hajer Kefi. ‘Instagram and YouTube Bloggers Promote It, Why Should I Buy? How Credibility and Parasocial Interaction Influence Purchase Intentions’. Journal of Retailing and Consumer Services 53 (March 2020): 101742. https://doi.org/10.1016/j.jretconser.2019.01.011.

Steers, M. L. N., Wickham, R. E., & Acitelli, L. K. (2014). Seeing everyone else's highlight reels: How Facebook usage is linked to depressive symptoms. Journal of Social and Clinical Psychology, 33(8), 701–731.

Petersen, A. H. (2019). How Millennials Became The Burnout Generation. AMASS, 23(4), 16+. https://link.gale.com/apps/doc/A592557454/AONE?u=anon~9b4b8065&sid=googleScholar&xid=e69ac96b

Mcintyre, Niamh. ‘Fossil Fuel Firms among Biggest Spenders on Google Ads That Look like Search Results’. The Guardian, 5 January 2022. https://www.theguardian.com/technology/2022/jan/05/fossil-fuel-firms-among-biggest-spenders-on-google-ads-that-look-like-search-results.

Watson, Gwendolyn Paige, Lauren D. Kistler, Baylor A. Graham, and Robert R. Sinclair. ‘Looking at the Gig Picture: Defining Gig Work and Explaining Profile Differences in Gig Workers’ Job Demands and Resources’. Group & Organization Management 46, no. 2 (April 2021): 327–61. https://doi.org/10.1177/1059601121996548.