Clipping is a red herring

Timeline control as a skillset

Bear with me for a somewhat more exploratory post today.

There seems to be a ton of chatter around “clipping” lately, especially on Tech Twitter. But clips are an incremental game. There’s a craft to it, but it’s not rocket science to turn longform source material into a shortform delivery vehicle. People are finding success with their own remixes on the format, like Dwarkesh’s special effects, but that edge on its own very quickly disappears if everyone else is doing it too. Substack even auto-generates clips for you now!

Clipping is just a small part of an emerging marketing skillset, which I’m loosely referring to as “timeline control.” I don’t have any answers here, but I think these are questions worth exploring.

New media properties win on emerging definitions of quality that they are structurally better-equipped to compete on

First, an aside on what people want from new media properties.

I’ve been stewing on this since my post on Olivia Unplugged, an edutainment creator who was incubated inside of the screentime app Opal. She’s closer to “Internet-Personality-as-Art” than “influencer.” To quote from my post at length:

Doug Shapiro has a fantastic post on the changing definitions of “quality” in media. In it, he posits that “new entrants don’t compete on existing measurements of performance,” instead they “introduce new attributes,” which “can change how consumers define quality.” Airbnb didn’t compete with hotels on 24 hour concierge services or high-end spas, instead they introduced entirely new attributes like a full working kitchen, or space to host guests. In the context of media, he cites a YouTube study where “55% of viewers agree that their content choices collectively create a sense of belonging.”

The degree to which you can be parasocial with a piece of media is a new attribute for its success. Olivia Unplugged, for example, is a person first, with all of the personality quirks and visual worldbuilding that Gen Z has grown to expect from shortform creators.

. . . .

When you contrast this against “New Media” in tech, especially launches or fundraise announcements, it’s clear that people are trying to apply old models of success to new formats.

…simply increasing production value or adding story structure to a Twitter-native video…is fundamentally missing the point of how media preferences can evolve.

It’s easy to look at Olivia Unplugged, realize that “in-house creators” are an interesting new format, and then copy it for yourself. I bet there’s some success to be had there. But the real competitive advantage is in understanding the emerging definitions of “quality” from your audience, and then designing a format to deliver on that.

Take TBPN, tech’s favourite new format of the year, which is effectively just a daily show livestreamed on Twitter. The format itself is not that hard to copy, which is why we’ve quickly seen fast followers like TITV and ETN. The copycats are flailing for the usual reasons that copycats fail, like a lack of real sauce. But they’re also flailing because the TBPN format is just a red herring — the copycats are not actually delivering on the new attributes of quality that people love TBPN for.

The new definition of quality that TBPN is delivering on is something akin to “creating a sense of belonging for Tech Twitter.” I often joke with my friends that people don’t even watch TBPN, everyone just loves that a “for-tech by-tech” media brand exists. Their daily livestream format is a structural advantage that helps them deliver on that “belonging” demand. Jordi and John, the hosts, live on the Twitter timeline, they’re actively shaping and replying to the discourse. So when they pull tweets up live on-air, or create in-group jokes that viewers can join in on, it makes a particular corner of Twitter feel connected. One of us… one of us…

Their format gives them a structural advantage on this new attribute of quality, but it is the attribute they are winning on, not the format. Competitors can copy the format, but they can’t win if they don’t understand the new definition of quality that the format enables.

Clipping is a tool, creating belonging is an objective

Clipping is also a red herring.

Internet anthropologists may disagree on the precise timeline, but clipping (and clip farming) is definitely not new. It was popularized by Andrew Tate in his TikTok takeover of 2022. Twitch streamers started using it around the same time, Tabs Chocolate adopted it for ecommerce in 2022, and then Jenni AI brought a version of it to consumer software in 2023.

I suspect tech’s recent hard-on over clips is once again because of TBPN. A cursory glance will show relatively low viewership on their YouTube channel and Twitter live streams, especially when you compare it to their outsized cultural impact. So when you look at your Tech Twitter timeline and see TBPN videos plastered everywhere, in the same way that copycats may look at their success and copy their livestream format, it’s equally easy as a viewer to attribute their success to their clips.

I think their clips are great, and work especially well with the bite sized interviews in their livestream format, but it’s once again missing the forest for the trees. Outside of being great hosts and thinkers, the artform that they’re actually good at is timeline control.

Timeline control as a skill

Admittedly this is where my thinking gets less clear.

So: we now know that “belonging” is a new attribute of quality, and format and clips are just tools to satisfy that demand. I think the skillset to satisfy that demand is something like “timeline control.” The ability to orchestrate a sequence of moments, aesthetics, and micro-events that help people feel something in particular.

If you think that the most important skill is in designing a clip or a good tweet, then there’s only so far you can push the frontiers of that craft. But if you zoom out to the craft of “timeline control,” there is a far wider aperture of skills to improve. TBPN, for example, is masterful at positioning themselves in the center of Tech Twitter’s ongoing conversation. Outside of clips and formats, they’re good at: “reading the waves” (knowing when a topic is peaking or dying), “owning concepts” (coining phrases, creating shareable aesthetics), and “network choreography” (how to get the right people involved at the right time).

Sure, this could broadly just be interpreted as “marketing,” but it feels unique in a few ways:

• Timeline control has to happen in real-time, in response to an actively evolving feed of information. This is naturally a better fit on Twitter than other platforms. You are always engaging with a meta conversation.

• Timeline control should seem decentralized.

• Timeline control shapes your sense of reality. Everyone is talking about this, so it must be worth exploring.

Two examples outside of TBPN:

1. OpenAI took over the internet earlier in March when they launched their image generation model, and everyone simultaneously “realized” they could turn pictures of themselves into Miyazaki’s distinct art style. I could have sworn I saw a quote of Sam Altman saying that they chose the Ghibli aesthetic on purpose to have a viral launch moment, but I can’t find it anywhere now. But it doesn’t take a leap of the imagination to see how all of this would have been a coordinated marketing campaign. The team chooses an aesthetic that they know will spread, they seed it across employees and key opinion leaders, and then you manufacture a moment on the timeline. It all feels a decentralized moment of connection between you and your corner of the internet, but it’s meticulously controlled.

2. A week before the Ghiblification of everything, my friends at New went megaviral with a satirical AI model. The week leading up to the launch, InternetVin started drama on the timeline with the cofounders of Cohere, a Canadian AI lab. He planted the seeds of a possible new AI lab, and repeatedly teased an incoming model over the course of a week. And then finally, at the peak of Indian hatred on the Twitter timeline, his team dropped an obviously fake AI model named “Brampton,” named after a city in the GTA with a predominately Indian population. “Brampton is very intelligent.” With the foresight of a few bubbling trends and a grassroots conversation to seed their launch beforehand, the timeline was putty in their hands.

I often think about a narrative device that Alan Moore used in the Watchmen comic. In the background of the main story arc, you’ll catch subtle references to various artists and writers disappearing — a newspaper header here, a TV broadcast in the background of a scene there. Nothing explicitly indicates that this is an important storyline, so it just feels like background context to make his world feel more alive. Slowly and then all at once, the final climax arrives and it turns out everything was connected to the disappearances all along.

“Timeline control” feels similar. It’s using a range of tools to exert subtle control over where people look, how they feel, what they think about, so that you can ultimately bring them into your universe. “Propaganda” has so much context that it feels like an inaccurate term here, but these decentralized marketing tactics all similarly assault our sense of reality: the hordes of UGC creators making videos for you, fake fan accounts, brand-run meme pages, or coordinated messages from “key opinion leaders.”

I think there’s more to explore here, and I don’t know nearly enough about it yet. I’d love to speak to whoever is the best in the world at it. Hopefully more to come on this front in the New Year :)

---

Every year I end the year with a few weeks totally off the grid. It is very cute when my friends know to email me if they need me after ~Dec 20th. You won't see a post from me next week, but I hope my loyal Attention Heads have a lovely Christmas and I’ll be back with a year recap the last week of December.