In my last post, I shared some takes about algorithms as a way of ordering a discourse. Michel Foucault talks about “discourse” as the ensemble of technologies, rituals, institutions, and rules involved in communicating ideas—and our discourses are always changing across history, just like the ways for making clothes or preparing food are always changing. (Every time I type the word “discourse,” I think of Twitter discourse, but that is not what I am talking about here.)
My idea of what’s happening with the election, or what happened in World War II, is as much based on facts as it is on the discourse I use to receive and understand those facts. That discourse is made up of A) specific choices and tools made by others to develop those facts into narratives and deliver them to us, and B) my specific situation and intentions when I receive the information. Algorithmic technology has changed the way we do this by involving automated computer processes at each step, which changes how the information gets formed and delivered. Information did not just fall out of a coconut tree, as Vice President Harris would say, it exists in the context of all which came before.
To make this argument less abstract and figure out more fully what I mean here, I’m going to describe in the posts that follow memes which, in their trajectory and patterns of audience reception, illustrate how a discourse ordered by algorithms works. Memes are canaries in the coal mine of the internet, digital artifacts whose journeys along the circuits of media discourse can tell us about how it all works. First, my primary area of expertise, Distracted Boyfriend, and one corner of that question: the relationship between Twitter and SEO journalism, which in 2017 was (and in 2024 kinda still is, to a lesser extent) fundamental to how information moves online.
Distracted Boyfriend
For my Master’s thesis on Distracted Boyfriend, I focused on the week of the meme’s virality in late August of 2017. Although DBF circulated prior to August 2017, it was this August 19th post by nm161 on Twitter, a since-deleted account, which lit the spark for a viral moment:
Observing memes as an editor at Know Your Meme, I’ve noticed how they tend to follow particular patterns of viral spread. I think this is best summed up in the chart below, posted to the MemeEconomy subreddit by a user named bogmire in 2017. On MemeEconomy, the organizing metaphor is memes = stocks. Valuation on the Y axis could be seen as overall number of posts and attention. I don’t believe all (or even most) memes follow the curve drawn here exactly, but I think it touches on something true. I wanted to better understand this pattern, which is why I focused just on that week of virality.
We can put January to August 2017 down as the “Stealth phase” (Meme Documentation on Tumblr discovered the earliest DBF posts dating back to January 2017 on Facebook)
The “awareness phase” is that weekend of August 19th-21st, right after that first viral posting. “Mania phase,” for me, begins around August 21st and lasts through August 25th, 2017. The 22nd, that Tuesday, was the “know your meme article written” date.
I gathered up around thirty posts about DBF written during the “mania phase” by different sites such as Buzzfeed, Daily Dot, and Digg as well as less-online publications like Wired, Le Monde, and The New York Times. August 24th (a Thursday) was when most of these posts came out, and then there was another crop of articles the next week (around August 28th, after the weekend) when the original stock photographer replied to a group of reporters with his comment on the story.
The most feverish period of posting of Distracted Boyfriend memes appears to have been between August 21st and August 24th, and so the internet culture reporters came in right around the “Buzzfeed” and “first panic sell” moment. I sread through these stories and found many of them follow a similar SEO-optimized pattern: four to seven brief paragraphs at the top, using keywords that will attract Google’s algorithm and place the article up high in search results, followed by a listicle-style presentation of different meme examples.
The same twenty to thirty Distracted Boyfriend memes also showed up over and over again in these listicles, regardless of site. The vast majority of these memes were sourced from Twitter rather than Reddit, Tumblr, Facebook, 4chan, iFunny, or any other big meme platform in 2017. Physically, they were embeds from Twitter in most cases.
When I interviewed the journalists who wrote these articles, they largely told me the same thing: they found DBF on Twitter organically on their scrolls, saw it was getting attention, and raced to write a quick piece about it, optimized for SEO. I understand the game because this is what I do nearly every day. Sometimes the titles involve “distracted boyfriend” or “man looking at other woman,” and they always involve “meme” because the goal is to get Google to serve up your article when someone searches for “meme with the guy looking at the other girl”— which, from about August 21-24, 2017 was enough people for you to get a good number of page views.
I draw from these observations of Distracted Boyfriend coverage a few take-aways about the order of “discourse” which moved information around online on August 2017, particular patterns in how information moved on that era of the internet:
Twitter was the link between the online world and the “real world.” This holds true for memes like Distracted Boyfriend but also for politics and posting of all kinds (e.g., Donald Trump). If it looks like people are talking a lot about something on Twitter, it’s usually worth writing a story about. I would argue that around 60% of all culture war stuff starts from reporters reading Twitter and reporting on it. Newspapers will often say stuff like “we’re the first draft of history,” but in the mid-2010s that changed: Twitter became the first draft, and they became the second.
The incentive structure of Google and SEO prioritized speed and specificity over all else. People had to write quick in order to beat other sites to the prize and get a good ranking. But they also had to write to highly specific prompts and keywords: internet journalism was never about broadness or getting a bird’s eye view but about the “deep dive” because the algorithm was powerful due to its ability to find exactly what someone was looking for, so you had to be exact. It created a journalism made of niches rather than beats.
There were a lot of people in this space, but only some of them got paid. A meme has value because it generates page views, clicks, attention, etc. I’m not sure exactly what the price tag on “Distracted Boyfriend in August of 2107” would be in terms of monetary value, but I’d estimate that SEO journalists/publications got maybe ten percent of that pie, influencers with some form of brand deal or sponsorship that posted the meme got five percent, and the platforms got eighty-five percent.1 All discourses have some material, economic basis to them, and in 2017 online discourse was subsidized by the free labor of a large group of people. Users were aware of this. Here’s the comments on one of the more viral Twitter postings of the meme, made by bruce bogbutter:
And here is the bogbutter meme in a Buzzfeed article, as an embed. I contacted bruce bogbutter, and was informed that they did change their profile name to something unhinged in order to screw with Buzzfeed, but that at some point Twitter changed the way embed codes worked so they were no longer live, and you couldn’t smuggle curse words into Buzzfeed.
It is flash-in-the-pan oriented. Most memes that pop off on Twitter and get hugely viral follow the same pattern as Distracted Boyfriend: two to three days of steady posting, two to three days of wild posting, and then a decrescendo of another two to three days after. Often, the steady posting will be over a weekend and the wild posting over a workweek. A consequence of this is that lore and narrative do not develop in the same way. Rather than a series of logically-connected developments that form a coherent chain (like, say the evolution of Wojak or Pepe on Reddit/4chan) you have a bunch of disconnected moments. This was the same pattern followed by controversies over Trump tweets, back in the late 2010s.
Each of these conditions impacts how discourse moves and the kind of information and opinions we get. At each step along the way, we encounter forms of resistance (the bogbutter Buzzfeed post being, for me, among the more evocative). The end result is a collaboration between human actors and the computerized algorithms they are trying to game, in which each primes, manipulates, and pushes the others to act. Motives vary from profit to ideology to just plain fun.
But maybe the most important lesson of it is this: Distracted Boyfriend could not happen in 2024.
Google and SEO no longer work. Half of the sites I looked at from 2017 aren’t even publishing articles anymore. The platforms have squeezed journalism out and taken the money — and whatever power journalists and media outlets had over the discourse, whatever check they held on the power of platforms or organized groups of users, is greatly diminished.
Twitter/X is also no longer the number one platform for meme origins. The discourse has moved to video-first, audio-second, and text/image-third. But it still incentivizes speed and specificity. Everybody has to “niche down” if they create content, and they have to work rapidly. This impacts the type and tempo of information we receive.
If anyone can think of a way to measure the monetary value of a meme, would love to hear about it.