Platforms end up converging into the same space with the same features

Interview with Zoë Glatt from Microsoft Research New England

I had an opportunity of chatting with Zoë Glatt, aka Dr. YouTube. She is a true internet culture aficionado, co-founder of The Digital Ethnography Collective and a postdoctoral researcher at Microsoft Research New England. Her research is an absolute goldmine of insights and inspiration for my research, and I could have talked with her for hours. The result of our conversation is the interview that you can dive into below. We discussed influencer industry, generative AI and digital ethnography.

You recently finished your PhD (🎉) in which you studied creator labor and precarity in London influencer industry. If you are not too sick of talking about your it, could you please tell me what are the main findings of your dissertation?

I’m not sick of it, actually it’s interesting for me to talk about it because I finished it in March and I’m preparing a book proposal 'Demonetised: Inequality, Co-option and Resistance in the Influencer Industry', based on that so it’s good for me to revisit it. ☺

My PhD was an ethnography of the influencer industry in London and I did different types of data collection: visiting industry events, online participant observation across platforms, I interviewed 30 creators and one social media marketing person and I also became a YouTuber for a year as a form of autoethnography.

The aim was to examine the nature of platformized creative work and I intentionally chose to look at creators across genres, platforms, levels of professionalization, as well as across identity categories like race, gender and sexuality, to get as broad a view of the industry as possible.

This was very much an inductive ethnographic project so I did not have concrete research questions at the beginning; they grew out of the research. The overarching research question that developed in the thesis was: How is creative labor being reshaped by the platformization of the cultural and creative industries? I draw on literature about the neoliberal worker subject in pre-platformized creative industries that argues that the promise of creative and passionate work leads to difficult working conditions and self-exploitation. Two sub-questions developed throughout the research: What are the distinctive sociocultural, technological and commercial structures that shape the experiences of content creators working in the burgeoning influencer industry? Second, Which creators are able to gain visibility and success, and how are certain groups and types of content systematically excluded in the influencer industry?

There are common myths of freedom, diversity and meritocracy in the influencer industry: that everyone can become a creator and be free to express themself. But what I found in my research was that not only do content creators have to contend with many of the adverse working conditions found in legacy industries, but they are also faced with new structural issues that intensify precarity and inequality, such as algorithmic recommendation and new organisations of power centered around the interests of advertisers. I found that there are actually many gatekeepers - some of them human and some of them non-human – that mean that only certain people and genres of content can be made visible and achieve success.

In my work I look at how exclusions are perpetuated at the commercial, technological and sociocultural levels. In attending industry events, I was able to witness things that happen in person and types of cultural dynamics between influencers and intermediaries like influencer marketing practitioners and talent agents. In general I found that the influencer industry promotes and is most comfortable with white, middle-class, uncontroversial creators who are “brand friendly” and therefore easily legible to sponsors and platforms. These influencers all have a certain types of cultural capital, which means that it’s easy for them to communicate with industry people. The commercial, the sociocultural, the technological structures interact with each other to cause multiple and compounding forms of exclusions for creators who are from marginalized identities and who produce content that is less brandable, for example those who are highly political or talk about taboo topics such as sex and sexuality.

Which brings us to the concept of algorithmic discrimination that you proposed in one of your very interesting papers and in the PhD thesis. Could you please describe it for my readers?

I define algorithmic discrimination as a process whereby certain content, identities and positionalities within the platform economy are deprioritized from recommendation in an industry where visibility is a key to success. Platforms have their Terms of Service and Community Guidelines that inform algorithmic governance that allows certain types of content and suppresses others through mechanisms of algorithmic sorting, shadow banning and demonetization, predominantly based on what will keep advertisers happy.

I draw heavily on Safiya Noble's work in Algorithms of Oppression, who talks about how algorithms are not neutral technological systems, but rather are trained by humans on data that is produced by humans, and so racism, sexism and other forms of inequality that structure society are reflected in these systems. I look at how this same dynamic informs the influencer industry and creator labour, alongside other scholars like like Brooke Duffy, Sophie Bishop and Carolina Are.

This dynamic is made more challenging for creators due to the perception that platforms’ algorithms are “neutral” non-human actors. It’s hard to creators to know what is going on if their content isn’t doing well and their income goes down, to understand why and even if they have been demonetised or shadow banned. And even if they know that something is going wrong they are unable to contact a human to challenge unfair discrimintation unless they are famous enough that platforms care about them being unhappy.

In your most recent article, you also demonstrate that creators from marginalized identities also face challenges besides algorithmic discrimination.  You proposed the notion of “intimacy triple bind” which captures how creators from marginalized identities or content genres are already at higher risk of trolling and harassment which puts them under increased pressure to perform relational labor, which in turn potentially brings further harms in the form of weaponized intimacy.

In my most recent article I wrote about how on top of systemic exclusion from income generating opportunities that rely on being brand friendly, many marginalised and stigmatized creators have to deal with much higher levels of trolling and harassment from audience members. The combination of these factors leaves marginalised creators in what I call the ‘intimacy triple bind’: they often need to rely on audiences directly for financial support and therefore have to perform what Nancy Baym calls “relational labour”, sharing personal aspects of their lives and beliefs in order to cultivate intimate connections with their audiences. But this relational labour adversely puts them at higher risk of what I call “weaponised intimacy” from hostile audiences looking to tear them down. So they are stuck between a rock and a hard place.

What I’m interested in exploring in my book is how platforms play a part in exacerbating all of these negative working conditions for marginalised creators while at the same time co-opting social justice narratives and positioning themselves publicly as caring deeply about inequality. You have initiatives like #YouTubeBlackVoices or Black Creator Fund and campaigns around LGBTQ+ creators. Platforms make statements about how they support marginalised creators and employ the language of anti-racism and feminism. But these are basically just marketing and PR activities because they are not invested in creating meaningful change in the industry that would fundamentally improve the lives of marginalised creators.

The bottom line in that in a commercial industry so intimately intertwined with the interests of advertisers, the profit motive always wins out, and meaningful anti-racism/feminism are the opposite of profitable. I wrote a chapter with Sarah Banet-Weiser about the branding of intersectionality, and how following George Floyd’s murder, when anti-racist and intersectional politics moved from the margins and into mainstream discourse, all companies felt the need to brand themselves publicly as anti-racist. We argued that this was corporate reputational management rather than a meaningful challenge to structural relations of power. Similarly, when we see platforms proclaiming their investment in social justice, it’s more about the spectacular display of positive ideas around diversity, rather than anything that would actually hurt the bottom lines of platforms.

Do you have any practical solutions for addressing the precarious and unfair working conditions of creators?

There are different layers of answers to this question because there are many different stakeholders: creators, platforms, brands, managers, digital media marketers and regulators. When thinking about what could improve working conditions, there are different answers depending on who you’re asking.

One answer is creator-led, which means unionization or other forms of collective action. Victoria O’Meara speaks about Instagram engagement pods as a micro-form of resistance in which creators like and comment each other’s posts to help each other gain visibility. It is not exactly resistance because it's just playing within the rules of the visibility game, but it is a form of collective action.

You have also various different examples of creators trying to collectivize more formally. The most visible version was the Internet Creators Guild that was started by Hank Green who is a central figure in the US creator industry. The Guild existed between 2016 and 2019 and had meetings with YouTube to give recommendations. They probably did have some impact on how the platform was designed but ultimately failed because the model did not work. The majority of their members were small, aspiring creators and the large famous creators did not really have an incentive to become members.

There are other examples like American Influencer Council which my colleague Sarah Edwards is writing an article about. They are much more commercially oriented and situate themselves as an intermediary between creators and brands. They bounded themselves specifically to the US creators which is kind of fucked up but it also makes sense because it’s more coherent. There are also UK and European examples. One initiative was called FairTube campaign and they were associated with the biggest trade union in Germany. Nonetheless, there are limits to unionisation. Niebler and Kern speak about technological, geographical, and organizational fragmentations that make it difficult for collective action or regulation to occur in this sector. 

Then there are also more top-down initiatives like the UK Parliamentary inquiry into influencer culture led by Sophie Bishop, which was designed to make recommendations to change the laws and regulate platforms, though regulation comes with its own unique set of challenges.

Your PhD research took place between 2017 and 2023 and considering the rapid pace of change, did you observe any significant changes in the creator/influencer industry that happened over time?

Already on week one, my supervisor Sonia Livingstone asked me: What if YouTube closes down in the next five years? With this question, she was trying to encourage me to ask questions that were bigger than thinking about single platforms. At the time I was like I don’t know!! But over time the project developed to be more about the platformization of creative labour and the cross-platform nature of the creator economy more broadly.

Zoë with creators (Jazza John on the left and Taha Khan in the middle) during her fieldwork in Riga, Latvia for the UNESCO Media Information and Literacy Week in October 2018

An obvious example of something that changed whilst I was doing fieldwork was that TikTok appeared in the West somewhere around 2018. When you go to the industry events, they are always three steps ahead talking about trends and where things are going. In 2018, I went to the VidCon and there were a lot of panels about TikTok.

New platforms rise and fall all the time. YouTube and Instagram have been consistently major platforms for the whole time during my research but TikTok just popped out of nowhere and exploded. It has changed the type of content that people are interested in creating and consuming. We’ve moved in a different direction where the popularity of short form content on TikTok has impacted the features of other platforms like Instagram reels and YouTube shorts. Me and Ysabel Gerrard from the University of Sheffield have been thinking about how different platforms compete to be the one-stop-shop ultimate platform that everyone will use and how platforms end up converging into the same space with the same features.

Predicting the future can be challenging, but I'm curious to hear your thoughts on the potential direction of this industry.

There's a lot of hype currently in the industry around the potential impact that AI tools will have. AI tools are being sold to creators as like “This will make your job so much easier! They'll edit your videos for you! You can make a website without having to do the work!” etc.

I don't really know how much uptake these tools have at the moment, but I think it will be very interesting to see whether the hype around how they're being sold matches up to the realities of what actually ends up happening and also whether, as we know with algorithms, these tools end up perpetuating the same inequalities and discriminations.

The ways that these things are being framed by Silicon Valley bros is very much like, “it's so cool, you can unleash your creativity, you won't have to do the boring stuff”. I have just started a project about AI in the creative economy called “10x Yourself: The promises and perils of generative AI in the creator economy” based on a creator training course being offered that claims that these tools are going to allow people to create 10 times more content and be more creative because they won't be worrying about doing X, Y, and Z because these tools will do it for them.

With respect to the future, at the end of every interview I did with creators I asked them what they thought the future of the industry would look like. Most of them were like, oh, it's a fool's game to try and guess what will happen. People had ideas about certain platforms at certain times, whether they'd be big or not. A big one was whether and if YouTube will die at some point. That that would be major for a lot of creators, because even though platforms like TikTok and Instagram are a bit more sexy, YouTube continues to be the main platform that offers any significant amount of advertising revenue share with creators.

Do you have any numbers of how much creators actually get from AdSense?

Not really and there's a reason for that. Platforms always make things very opaque and this lack of transparency is quite a big thing. There are initiatives that try to make brand deals more transparent, like F*** you pay me, but the same doesn’t exist for AdSense.

I spoke to one sex education creator and has about 800,000 subscribers, which is a lot for that vertical. She has two YouTube channels. Her main channel (around 700k subscribers) is about sex education and is less brand and family friendly. She found that her videos were always being demonetised so she started her second channel (around 80k subsribers), where she focuses on lifestyle, talks about her child and other uncontroversial topics. When I spoke to her about this she had had 429k views that month on her main channel 183k views on her second channel, but had earnt a third more AdSense revenue on her second channel, due to its higher CPM (cost per mille) and lower levels of demonetization. Crazy!

That is just like one example, obviously, of one particular type of content, but there are so many different calculations that are going into how much AdSense revenue someone earns, and it is very complicated. But I’m very interested in how CPM calculations once again privilege certain types of content and creators over others. For example, if most of your viewers are based in America and if you make tech reviews you have a very high CPM, which heavily privileges white, Western men.

You are also co-founder of The Digital Ethnography Collective. How would you describe digital ethnography?

So, to give you the long short story, I started out in Anthropology and I wanted to do my BA dissertation about a particular YouTube community, but it was difficult because anthropology is quite resistant to online forms of ethnography because it's so foundational for the whole discipline, going to a place and living there and being in it.

The reality is that doing digital ethnography, like when I say digital ethnography, in this case, I'm talking about specifically online participant observation, it introduces so many complexities to the nature of fieldwork and what it means. I don't think anyone is really saying digital ethnography is the same as more conventional ways of doing ethnography, except in the case of more traditional ethnographies that happen in digital, bounded digital spaces, like Coming of Age in Second Life where Tom Boellstorff did a traditional ethnography but in the virtual world of Second Life.

Most digital ethnographers would not go quite that far to say this is a world of its own because digital media is part of offline life and online life. In my PhD I decided to combine online and offline ethnography because it was clear to me that if I wanted to understand the lived experiences of content creators, I could not really just observe them online, because they also exist offline. And there’s many industry-level things happening offline. What you see online is one aspect of their labour, it's kind of the product of their labor. However, what is happening in their lives and how they interact with each other and how they interact with industry stakeholders, that all mostly happens offline. For my questions specifically, it made sense to look both online and offline.

Interesting thing is that I could not just study the influencer industry as one thing because it depends on where you are. Local norms inflect everything about how people use technology. So I chose London, but I added a major caveat in my methods chapter: these UK creators are existing on platforms that are global and they might interact with other people from other places and basically English language creators are all kind of in one industry in cultures that are produced internationally. So I ended up doing some fieldwork in the US as well because that’s where the biggest Video conference takes place and quite a few UK creators go there. That was one of tricky things, how do you bound the digital culture which is international?

Long story short, I ended up turning to literature outside of anthropology because I found that they were less rigid about the method. So for example, I've got my bookshelf here: Internet Inquiry by Annette Markham and Nancy Baym is super useful and they were so ahead of time, this books is from 2009. Also Christine Hine’s book Ethnography for the Internet was also super useful.

As a part of your data collection strategy, you also employed a unique method of autoethnography and became a youtuber for one year. Are you going to revive your YouTube channel?

I am happy to leave this stuff up. I even watched one of the vlogs the other day, a vlog from going to LA for fieldwork. It was just interesting to reflect on what I was looking at then, it is five years old now which is crazy. It's a lot of work to revive your channel that was the main findings that was useful to my PhD, just how much work it takes to run a YouTube channel. I just don’t have the bandwidth to do it justice.

---

🎨 Are there any influencers/creators who you would be sad to see stop posting?

Yes, a lot! I am subscribed to around 1000 YouTube channels, I’ve been super invested in the platform since I made vlogs back in 2008. Some of the people I really enjoy watching are people I ended up interviewing. I really enjoy watching:

Tiffany Ferg: https://www.youtube.com/@tiffanyferg

Khadija Mbowe: https://www.youtube.com/@KhadijaMbowe/videos

Frannerd: https://www.youtube.com/@FranMeneses/videos

Ariel Bissett: https://www.youtube.com/@ArielBissett

Micarah Tewers: https://www.youtube.com/@MicarahTewers

Rachel Maksy: https://www.youtube.com/@rachelmaksy

Jarvis Johnson: https://www.youtube.com/@jarvis

Leena Norms: https://www.youtube.com/@leenanorms

Hannah Witton: https://www.youtube.com/@hannahwitton

Some of them are quite well known and some of them are a bit smaller. There are still so many people doing great stuff on YouTube. I was really sad when Jenna Marbles stopped posting. Often in the comments section on TikTok someone will say “Jenna would love this”. It's so fascinating to me that she still has that level of hold over people's imaginaries. Obviously she went through a lot of different phases in her career and some of her earlier stuff was in poor taste but I think she really grew up and transformed over time. There was a lot of joy on her channel, something that many different people could engage with and enjoy.

📚 Which books would you recommend reading?

In addition to all of the methods books, I have five books here. Number one: (Not) Getting paid to do what you love by Brooke Duffy. It's just such a useful book when you're thinking about labor, gender, influencers, it’s really good.

Another kind of obvious one is Creator Culture. I am biased because I have a chapter in it with Sarah Banet-Weiser, but it is a really good book. The foreword by Nancy Baym is super useful.

Probably the most useful I found for my PhD is The Intersectional Internet: Race, Sex, Class and Culture Online. It's an edited volume by Safiya Noble and Brendesha Tynes. They proposed a new field of study called Intersectional Critical Race Technology Studies. It's basically an intersectional feminist framework of thinking about how the internet structures and reflects culture in terms of inequalities and identity categories. This is a bit of a classic for me.

And then a textbook, the one that I used the most in my thesis was The Cultural Industries by David Hesmondhalgh. It has just everything about debates around cultural and creative industries!

Techno-optimist or techno-pessimist?

I think in my work I am a techno pessimist, because that’s often the nature of being a feminist scholar, unfortunately. But actually, I was encouraged to rethink this the other day. A group of scholars came to visit here who were doing a project on race and technology, and they were talking about literature around black joy and optimistic frameworks when it comes to critical scholarship. It made me think twice because it is very easy to replicate negative kind of pessimistic accounts in this field. Therefore, the reality is at the moment, I am bit of a pessimist in my work but an optimist as a person. I'm an optimist in a sense that I get a lot of joy out of like watching content creators, understanding what they like about what they do, and trying to think about how that joy could make it’s way into my work!