It's Time to Call Top on Startup Hype Videos
A startup trend covered by the New York Times is its own type of obituary. But it's so much worse than that.
The New York Times published a piece this week about startup hype videos, the genre of highly produced, $80,000-and-up short films startups (and especially AI startups) commission to announce a funding round, a product, or... just vibes. It looks at Daydream, which built an entire Alice in Wonderland set, rented a $2,000 rabbit-head costume, and paid a 20-person crew for two days to announce a $15 million raise in a social media post.
It’s all but law that when the Times pulls up to write about a startup trend, the trend has peaked and begun its inevitable descent into cliché. So we can safely call the top here. The startup hype video is over.
I could not be happier about it.
Let’s start with a marketer quoted in the piece, who put it plainly: there are “50 start-ups working on the exact same thing, and often the differentiator comes down to marketing.” Marketing can be a differentiator in a category, but only if you’re employing different marketing strategies and tactics. When every startup is making one of these videos with the same studios and the same look and feel and beats, they become interchangeable and forgettable.1
But I’d go one further. If you’re a seed-stage company and marketing is your differentiator rather than your product, you fucked up. A company that’s fundamentally undifferentiated from the jump is already failing, whether or not anyone’s said it out loud yet. This is the rare thing Y Combinator co-founder Paul Graham and alleged-new-Argentinian Peter Thiel actually agree on. PG’s whole instruction amounts to “make something people want”; Thiel’s is that competition is for losers, that the only businesses worth building are the ones doing something no one else does.
Fifty companies doing the same thing at varying degrees of volume is the exact opposite.
The deeper problem is that these videos are a category error. At the early stage, your brand is your product. You don’t need brand marketing because the brand is supposed to be the thing you built and the people you built it for. Brand marketing (meaning, promise, mood, the stuff a hype video is made of) is what you graduate into later, once you’ve found product-market fit and others begin imitating your differentiation or you’ve accumulated enough lore (or aura, to use another Gen Z word) that it needs curation.
So when one of these videos shows up in my feed somewhere, it usually reads to me one of two ways: a startup spending a hefty chunk of change to chase clout over customers — after all, the primary consumers of these startup videos are... other startup founders and funders, both of whom are supportive but insular cliques — or a company whose product pull isn’t there, papering over it with atmosphere. Usually it’s a little of both.
But the thing in the piece that actually set me off wasn’t anything about the videos, but this quote from Aaron Epstein, a general partner at Y Combinator:
“If people knew it was just, you know, two people in a living room that hacked this thing together over the last few weeks, people might not trust it as much, and so it’s one way to build trust with the audience.”
Insert guy-blinking-in-disbelief meme here, but it’s my face.2
I’ll grant there may be context from the interview the quote strips out. But evaluating it purely on its face, this reads as short-term and self-defeating advice from an institution whose entire job is to tell founders to launch early, ship, and get real feedback from real users. Epstein isn’t an anomaly in the Valley; he just said the quiet part out loud.
A startup attempting to buy trust with a hype video is kind of like a trope of LA dating: guys who drive BMW M3s, but who live in a dump in Reseda. Sooner or later (usually sooner), you get found out. That’s not earning trust. It’s betraying it and salting the earth so it never grows again. You simply cannot growth-hack your way to trust-building, because doing so drives an equal and opposite reaction.3
Now, every startup runs on a little projection. But it’s usually aligned projection: radiating a belief that you understand a problem and the people who have it better than others, that you know exactly how to solve it, and are hell-bent on doing just that. That confidence is both real and a precondition. Founders who can’t project it don’t get far, and shouldn’t. But that’s confidence in your approach, not a disguise for what you are. The hype video is the disguise. Go back to what you actually have at this stage: nothing but the product doing something real. That is your trust.
None of this should be read as anti-video. I’ve argued comms teams should build video teams, and I’m betting on video myself this year. The line isn’t video or no video, but video that shows you something real versus video that exists so you don’t look behind the curtain.
Like any good trend that’s peaked, we already have the cautionary take, and it’s right in the piece too. Cluely reportedly spent $140,000 on a video of its CEO using AI to lie about his age on a first date, under the tagline “cheat on everything.” A couple of months later, it raised $15 million from Andreessen Horowitz. So the stunt “worked” in that sense.
Then actual users showed up, found a product that didn’t do much, and Cluely quietly swapped the tagline and repositioned as an AI meeting assistant and note taker. The spectacle bought a funding round, in the same way that M3 might get our proverbial LA ladies’ man a date. But it didn’t buy a business or a working product. Those were always two different audiences, and the campaign spoke to only one of them.
So no, I won’t be mourning, but I’ll stop by the funeral to make sure it’s dead on my way back to earning trust the only way it’s ever worked: the slow, unsexy way.
A sign of what an industrial complex this cottage industry has become: Nen Creative, the agency behind the Daydream video, told the Times it’s running an average of one to two shoots a day, five days a week.
The guy from the meme lives in San Francisco and seems like a pretty good dude.
A YouGov study published this month asked Americans what they want from AI companies. The clearest signal across responses: trust earned through proof, not hype. Go figure!