Y Combinator4 min read
Farza Majeed: founders should ship demo videos daily, not polished launch films
The Hey Clicky founder tells YC that scripted, phone-shot "demo" videos beat expensive launch productions, and explains how he measures which channels actually convert.
AI summary of “This Is How You Should Actually Make Launch Videos”
Key takeaways
- Farza Majeed says his viral Clicky demo got nearly 3 million views but only 7,000 downloads, and about 100 people retained after 5 days.
- He argues Twitter converts worst, while a 20,000-view YouTube video had roughly a 20% download conversion rate.
- Farza says every video is scripted days in advance but rehearsed until it sounds unscripted, calling audio quality the most important technical factor.
- His old Zip School launch videos got only 23 likes, which he still counts as a real "launch."
- He challenges founders to post one 10-minute phone video a day for 30 days to build the habit.
Farza argues "launches" should really be daily demos
Farza Majeed, founder of Hey Clicky and previously Buildspace, told YC's Aaron Epstein that the old model of saving up for one big, expensive launch is outdated. He said founders used to need a Hacker News post, a Product Hunt launch, or press coverage, but now "there are so many new channels now where our users are living." His view is that founders should treat every feature ship as a mini launch rather than waiting for one big moment.
He said he prefers calling these "demos" instead of launches:
"I'd rather call these demos instead of launches and I just love shipping just demos constantly."He described working backward from the demo when planning features, asking what the Instagram caption and hook will be before the feature is even built.
The Clicky launch video: nearly 3 million views, but downloads mattered more
Farza walked through the original Clicky launch video, filmed in the office five minutes before dinner. It reached nearly 3 million views, but he said he does not care about view counts:
"I actually don't care about this number at all. I was like, 'How many people downloaded it?'"By his account, about 7,000 people downloaded the app after that video, and roughly 100 of them kept using it daily for five days straight — the signal that convinced him to keep building what became his company.
Scripting videos that sound off-the-cuff
Farza said every video is scripted in advance, including the "screen-aware dictation" video, which he called
"the worst video we've ever done"before posting — it went on to get 3 million views on Twitter. He said he rehearses lines mentally for days before filming, and that the first 10 seconds of a one-minute video should cover who he is, what he's working on, and why it matters, since "most of these videos are 1-minute long" and don't have room for a long setup.
His test for whether a hook works is simple:
"Did my eyes stop?"He said he studies Instagram Reels and other viral formats for structural templates, not for entertainment, and reuses those beats in his own scripts.
Starting from zero: the Zip School videos with 23 likes
Farza showed an old 2021 Zip School video he called genuinely weak, which got 23 likes and one comment. He said that still counted as a launch in his mind:
"To me this is exactly what launching means... it doesn't matter if we got 23 views."He framed years of low-performing videos as the necessary repetition before he could make videos that worked, comparing it to needing to take a thousand basketball shots before making the five-thousandth.
Channel performance varies sharply by platform
Farza said Twitter has historically been his worst-converting channel despite high view counts: the Clicky video's 3 million Twitter views produced only 7,000 downloads, which he called a conversion rate "like less than 1%." By contrast, he said a YouTube video with only 20,000 views and 872 subscribers converted roughly 20% of viewers into downloads, calling YouTube "a library that will last for decades" compared with Twitter's one-time moment. He said Instagram drives around 70% of his company's traffic and is where his team tracks link-in-bio conversion most closely.
A real production stack built around a filmmaker
Farza described three tiers of gear: Screen Studio for screen recordings, an iPhone with Instagram's Edits app for casual selfie videos, and a Sony FX3 with a DJI mic for more cinematic pieces. He stressed audio quality above visual polish, saying he will stay for bad 720p footage if the audio is clean, but will "bounce" from good visuals with bad sound. He credited his filmmaker collaborator Vikrant with shooting, editing, and motion graphics on nearly all Hey Clicky videos, and said hiring a filmmaker was one of his first two company hires, alongside an engineer.
His challenge to founders: one video a day for 30 days
Farza closed with a direct challenge: spend 10 minutes a day recording what you worked on and post it, even to a brand-new account so existing followers don't see early attempts. He said that within 30 straight days, one video will likely surpass 1,000 views, giving founders a data point to learn from and improve on. He framed early cringe as a necessary stage:
"You have to be cringe for a little while actually to learn how to not be cringe."
Written by AI from the video's transcript. It can compress, misattribute or miss context — the original video is the source. Not investment advice.
