Farza Majeed Dumps Agency Launch Videos for Daily Smartphone Demos

Farza Majeed 弃用机构制作的发布视频，转向每日手机实机演示 初创团队创始人必须放弃昂贵的发布会排场，转而进行每日未经精修的实机演示；因为软件代码极易被克隆，而通过持续更新进展积累的受众才能筑起坚固的转化护城河。 Farza Majeed 认为，创始人应停止将产品发布视为高风险、昂贵的一次性大事件，而是应转向在 Instagram、X 和 YouTube 等算法平台上每日分享面对镜头的实机演示。通过在写代码前先反推分发噱头，并学会撰写听起来自然随性的脚本，创始人可以建立起比企业广告更能引起用户共鸣的真实叙事。他深入剖析了病毒式传播背后的指标，揭示了为何与精准的长尾转化相比，高浏览量往往只是虚荣指标——例如一段仅有 2 万次观看的 YouTube 演示，其转化效果却大幅超越了一条 300 万次曝光的推文。Farza 强调，讲故事和社区比软件更难被克隆，他还详细介绍了自己的分级制作配置，指出音频质量是决定成败的首要技术因素，并向创始人发起了一项 30 天挑战，鼓励他们克服对“丢人”的恐惧，坚持每天发布进展更新。 Farza Majeed 反对制作耗资 2 万美元的高精良度发布商业广告，主张在开始编写代码前先根据分发钩子进行反推，制作持续且日常的面对镜头实机演示。 (2:56) 表面的浏览量数据往往会误导创始人：一条拥有 300 万次曝光的 Hey Clicky 发布推文仅带来了约 7000 次下载（转化率不足 0.25%），而一段 2 万次观看的 YouTube 视频转化率却高达约 20%。 (31:13) Instagram 出人意料地贡献了 Hey Clicky 漏斗顶端约 70% 的入站流量，在驱动实际应用安装和合格线索方面的表现远超 Twitter。 (31:13) 在软件产品能被瞬间克隆的环境中，受众和分发能力成了主要护城河，这使得 Majeed 在早期除了招聘工程师之外，还优先招募了一名视频制作者。 (39:20)

Farza Majeed argues that founders should abandon expensive, single-day promotional spectacles in favor of daily, direct-to-camera demos that treat distribution hooks as integral to product development. Demos as Ongoing Narratives Over Hero Launches (2:56) Farza Majeed contrasts the legacy PR model of single-day press embargoes and $20,000 agency commercials with continuous, direct-to-camera demos. Because platform attention has transitioned from static aggregator directories like Hacker News and Product Hunt to algorithmic feeds on X, Instagram, TikTok, and YouTube, startups now launch in front of continuously refreshing audiences. Instead of coding features in isolation and devising marketing afterward, Farza works backward from the distribution hook, testing the caption and opening lines before engineering begins. The Illusion of Virality and Downstream Conversion (6:58) Top-line view metrics often fail to reflect actual product adoption. Farza notes that Hey Clicky's initial launch video generated nearly 3 million impressions but converted into roughly 7,000 downloads, representing a conversion rate under 0.25%. True product validation emerged only after tracking a core cohort of approximately 100 users who maintained a five-day usage streak, prompting direct customer interviews. Similarly, a YouTube demo on a channel with 872 subscribers attracted only 20,000 views yet achieved a ~20% download conversion rate, whereas a comparable X video with 2.9 million views yielded minimal account creation. Instagram generated roughly 70% of Hey Clicky's top-of-funnel inbound traffic, using dedicated bio links and desktop handoff flows. Scripting, Narrative Beats, and Scroll-Stopping Hooks (7:47) Despite an informal, direct presentation, Farza scripts launch videos in advance and reviews copy by injecting draft text into live feeds to test whether it halts scrolling within three seconds. Adapting standard cinematic structure, he dedicates the first 10% to 20% of a short clip to answering what is being built, why it matters, and immediately pivoting into the product. Pacing relies on narrative curiosity rather than hyperactive jump cuts. For example, Hey Clicky's hands-free computer control video ran in a single continuous take, progressing naturally from opening Spotify to checking a Stripe dashboard showing $140 in daily revenue. Rather than lifestyle vlogging, raw technical updates—such as bi-weekly standups detailing agent orchestration fixes—effectively engage early adopters. Early Video Investment and Technical Production Rules Because code can be cloned rapidly while authentic audience trust is difficult to replicate, Farza prioritized visual storytelling early, employing two video producers in a seven-person company at Buildspace and hiring a filmmaker as one of his first additions at Hey Clicky. For equipment, he designates distinct tiers: Screen Studio ($30) for desktop captures, smartphone selfie cameras for mobile updates, and Sony cinema bodies (FX3/FX30) with DaVinci Resolve for quarterly hero videos. Farza identifies clean audio as the single non-negotiable technical requirement, advising founders to bypass internal laptop microphones and AirPods in favor of dedicated wireless microphones. To overcome internal hesitation and peer judgment, Farza challenges founders to film and publish a 10-minute daily recap for 30 consecutive days, projecting that at least one clip will surpass 1,000 views to supply empirical baseline data on effective hooks.

Farza Majeed 与 Aaron Epstein 主张，初创企业应摒弃耗资数万美元、依赖传统公关的单次发布会模式，转向基于算法推荐、以手机录制的日常轻量演示（Demo）。他们认为在现代分发格局下，写代码的门槛已大幅降低，而通过真实持续的故事构建受众壁垒与获取留存用户才是增长的关键。 从单次发布会转向算法驱动的日常演示 (2:56) 过去初创企业依赖将产品提交至 Hacker News、Product Hunt，或等待 TechCrunch 与传统纸媒的单次专题报道，因此往往耗费数万美元制作精致商业短片。Farza 指出，传统媒体的把关人地位已减弱，用户注意力已全面转移至 X、LinkedIn、YouTube 和 Instagram 的推荐算法流中。这意味着创始人不再是发布单次公告，而是持续面向不断刷新的算法受众进行展示。创始人直接对着手机或电脑说话的粗颗粒度演示，在实际互动与转化上往往优于昂贵广告。 逆向构思钩子与脚本叙事节奏 (7:47) Farza 提出了以演示驱动开发（Demo-driven development）的工作流：在团队开始编写代码之前，先确定社交媒体文案和视频开头第一句话，以此在投入工程成本前验证概念吸引力。在文案测试中，他会将草稿通过浏览器检查元素工具粘贴到实时信息流中滚动滑过，自检视线是否会停驻；视频画面则在开局前 3 秒以界面特写抓住受众。即使视频看起来是随性录制，他也坚持提前撰写脚本并对照目标受众朗读，确保在最初 10% 到 20% 的时间（前十秒）内交代清楚在做什么以及为何重要，再无缝切入演示。 虚荣播放量与真实用户留存的转化落差 (6:58) 播放量属于虚荣指标，产品验证取决于漏斗深度的留存表现。Hey Clicky 的早期发布视频虽获得近 300 万次观看，但仅带来约 7,000 次下载，转化率不足 0.25%；而真正促使该项目成为全职业务的，是连续 5 天每日活跃使用的约 100 名核心用户。在不同渠道的表现中，Twitter 尽管浏览量大，但转化率极低；相比之下，只有 872 名订阅者的 YouTube 频道发布的一支 20,000 播放量视频，下载转化率却高达约 20%。此外，Instagram 贡献了 Hey Clicky 约 70% 的顶层应用流量，甚至面向 B2B 的设计机构客户也能通过精准定向的短视频获得线索，打破了短视频平台无法做企业级获客的固有偏见。 克服尴尬感与展示真实构建过程 (24:53) 很多创始人因担心熟人评判或自感不完美而不敢发布内容，Farza 认为忍受早期的笨拙（cringe）是必经阶段，受众渴望的是真实叙事而非包装过的企业公关。Hey Clicky 团队每两周发布一次长约十分钟录制的“本周所建一切”工程更新，没有夸张钩子，仅靠工程师坦诚沟通修复的问题与新进展，在 Instagram 上同样获得了近百万次观看。Farza 强调，这种内容展示的是日常构建而非创始人自恋式的日常生活 Vlog，早期软件企业完全可以通过让用户结识背后的工程师来建立信任。 影视化团队配置与音频质量优先法则 (38:48) Farza 认为代码复制极快，但受众社区和共鸣难以被竞品抄袭，因此 Hey Clicky 在创立之初便将早期招聘重心向视频倾斜，甚至团队一度拥有三名视频制作人员对应一名软件工程师。在硬件和工具选择上，他使用 30 美元的 Screen Studio 录制电脑演示，日常内容使用 iPhone 前置镜头，重大节点则使用索尼 FX 系列搭配 DaVinci Resolve。同时，他将音频列为视频绝对不可妥协的要素，强调观众可以忍受昏暗粗糙的画质，但糟糕的音质会导致受众立刻流失，因此创始人应坚决放弃笔记本内置麦克风或 AirPods，转而使用独立无线麦克风。 Farza 建议那些在私下闭门造车的开发者进行为期 30 天的发布挑战：每天花费至多十分钟，用手机前置摄像头记录当天攻克的技术或功能并发布至社交平台，甚至可以使用匿名账号。他指出，随着视频数量积累，至少会有一条内容突破千次播放，从而为创始人提供持续迭代分发能力的现实数据反馈。

[27:19] Farza Majeed: You got to be cringe for a little while actually to learn how to not be cringe. [35:20] Farza Majeed: I'm always just trying to keep the story moving. I'm not trying to keep your attention. That's actually the wrong way to look at it. [31:13] Farza Majeed: YouTube is a library that will last for decades. Twitter, you got your moment once and you're gone. [40:49] Farza Majeed: The most important thing in any video you make doesn't matter if it's a shitty Loom or a shitty like little video or your iPhone is audio quality. If your audio is bad, I'm bouncing.

[0:00] These days, too many founders think [0:02] launching is all about spending a ton of [0:04] time and money to produce [music] a [0:05] slick video and hoping it hits. But my [0:07] guest today, Farza Majeed, has a totally [0:10] different approach to launching and [0:12] going viral. [0:13] >> This is everything we shipped this week. [0:14] Well, shipping, right? [0:15] >> Are we shipping? [0:17] >> He's the founder of Hey Clicky, [0:19] >> [music] [0:19] >> where his posts regularly get millions [0:21] of views. [0:21] >> My formula is always the same. [0:23] This is who I am. This is what's going [0:25] on. This is what I'm working on. Let me [0:27] show you. Everyone wants a story, and so [0:29] how can a demo just be a mini story? [0:31] >> And today, he's going to share [music] [0:32] his playbook for how to create videos [0:35] that actually convert. [0:38] >> [music] [0:44] >> Farza, thanks for being here. [0:46] >> No problem. [0:47] >> Before we jump into some of your advice, [0:48] why don't you tell us a little bit about [0:49] your background, how you got to this [0:50] point? [0:51] >> Yeah, started my first company when I [0:52] was 13, selling stuff on eBay. 10 years [0:55] later, started getting into software and [0:57] built a online school for kids, which I [0:59] did with YC in 2020 with you. [1:01] That turned into a company called [1:03] Buildspace, which became one of the [1:04] largest schools for people who wanted to [1:05] work on their own ideas, and now working [1:07] on Hey Clicky, a new interface for AI. [1:10] >> Doing videos has always been kind of a [1:11] core strategy for you for how to reach [1:13] out to people, and especially with [1:15] Buildspace and now with Hey Clicky, it's [1:18] how you've gotten a lot of attention for [1:20] for your your products and your [1:21] businesses. How have you thought about [1:23] that and talk about your progression [1:24] from when you got started with video up [1:26] until now? [1:27] >> Yeah, I mean, to this day, I just love [1:29] making videos, making online videos for [1:30] people. And these days, I'm working on [1:32] my own company, and so the story that's [1:34] in front of me is actually my company's [1:36] story. And so it's been really fun to [1:37] just tell that online. And it just so [1:40] happens that making videos in 2026 is [1:42] really good for business, too. So, yeah. [1:44] >> Yeah, tell us about that. So, you've [1:45] been making videos for over a decade, [1:48] publishing them online. Talk about [1:49] what's changed. Like, what is different [1:51] now in making launch videos and getting [1:54] attention for your product in 2026 [1:56] that's different from how it was over [1:58] the last decade? [1:58] >> Yeah, I mean, you know, we I mean, you [2:00] were in YC together in 2020. You know, [2:02] you as my partner. I mean, yeah, back [2:03] then, you know, you were doing you had [2:04] to do stuff like post a Hacker News, [2:06] post a Product Hunt, go to Reddit, like [2:08] find the right subreddit. You know, like [2:10] that those types of launches. And [2:13] things have changed so much now. There [2:15] are so many new channels now where our [2:16] users are living across all all [2:18] products. It doesn't matter what you're [2:19] working on. Could be drones or AI or [2:21] like whatever a dating app. Your users [2:24] are on many new platforms. Mostly [2:25] algorithmic platforms like X, LinkedIn, [2:27] Twitter, YouTube. The biggest change is [2:29] you're not really making posts anymore. [2:32] You're actually just constantly [2:33] launching all the time on these on these [2:35] algorithmic platforms. [2:36] >> Yeah, we see that a lot with founders in [2:39] the YC batch where they think a lot [2:40] about they have to put in all this time [2:42] and energy and money to make some big [2:45] fancy launch video and coordinate this [2:47] big coordinated launch in the hopes that [2:49] that's going to be their big moment and [2:51] everything's going to take off. But you [2:52] don't think about it like that. Like I [2:54] think your perspective is very [2:55] interesting on this. [2:55] >> Yeah, there was a time where that was [2:57] the the meta. You know, like you got to [2:59] do like a fancy thing and make it a big [3:00] deal cuz you got one shot. You get that [3:02] TechCrunch article or you get like that [3:04] the press in the New York Times or [3:05] something like that. You needed [3:07] something kind of nice to go with it. [3:08] But now all the old media is kind of [3:10] just gone. And so now what's working is [3:12] like a video where I put I say hi and I [3:14] talk about my product is probably going [3:16] to perform way better than like a [3:18] $20,000 produced video by the New York [3:19] that I posted with the New York Times in [3:22] terms of actual engagement and traffic [3:23] and like other things. And so it's a [3:25] crazy world now. [3:26] You don't need any of old media stuff [3:28] anymore. And if you have a phone and a [3:29] laptop, you're so much more powerful. [3:31] And so yeah, the I think that now [3:34] we keep calling these launches and I [3:36] think that that is what these are. These [3:38] are mini launches all the time. But in [3:40] my mind, I never consider it I do [3:41] consider it a launch, but not in the [3:43] sense of like this is a big launch. This [3:45] is a big thing. It's not like we're [3:46] launching a rocket, you know. It's [3:47] funny. I'd rather call these demos [3:49] instead of launches and I just love [3:50] shipping just demos constantly. [3:52] >> Yep. [3:52] >> And the more demos we can put out as a [3:54] company, the better. At my company, I'm [3:56] always like, well, what are we launching [3:57] this week? That's more of the question. [3:59] And we work backwards always from the [4:01] launch of whatever we're working on. So [4:02] for example, I'll give you a small [4:03] example. If we have a feature we're [4:05] working on, well, we talk about it, sure [4:06] we talk about the feature, how it feels, [4:07] etc. But, then I ask, "Well, what's the [4:09] copy on the Instagram caption? What's [4:11] the first What's the hook? Uh if we put [4:13] If we make this a a video about this?" [4:15] And not that like it needs to have a [4:16] good needs to go viral, but we should [4:18] talk about it because how we communicate [4:19] to other people is just as important as [4:21] what we're building. And so, I'm it's at [4:23] every part of the process. It's not like [4:24] we make it and then magically I make a [4:26] hit video or something. That's never how [4:27] it works. It's always like we kind of [4:29] know it's a hit before we even start [4:31] like working on it cuz it just sounds [4:33] good on paper. [4:33] >> Yeah. [4:34] >> And so, I can show some examples of [4:35] that, but [4:35] >> I I think a lot of people um see videos [4:38] that you publish that go viral and get a [4:40] ton of attention, and they almost think [4:42] that it happens by accident because you [4:43] just seem, you know, so comfortable and [4:45] natural and off-the-cuff, but that's not [4:47] how you do it at all because you're [4:48] trying to figure out the hook before you [4:50] even do the video. So, maybe um it'd be [4:52] interesting if you could pull up a [4:53] video, walk us through it, and then [4:55] maybe walk us through the process of how [4:57] you created that video, what goes into [4:59] it before you've even actually started [5:01] shooting anything. [5:02] >> Totally, totally. My [5:03] uh philosophy on a lot of the launch [5:05] stuff now is just [5:07] how can we just show [5:09] like almost like the process of what I'm [5:12] doing versus show the ending of what I'm [5:14] doing. Uh I think there's a lot of value [5:16] to being like, "Hey, I'm this founder [5:18] person and I just got this much money [5:20] and this is my product and this this is [5:21] my fancy DSLR and these are my good my [5:23] my wonderfully edited video, you know, [5:25] with me in front of a nice office." You [5:27] know, but I'm sure we've all seen those. [5:28] You see those every day. I saw like five [5:30] today for sure. [5:31] >> Yeah. [5:31] >> And those have a purpose, but to to me [5:34] that becoming like the kind of like [5:37] thing that people are reaching towards [5:38] is really annoying. And so, uh I try to [5:40] always do the opposite and just do what [5:42] I think is right cuz again, that's a [5:43] one-time thing. You want You want a [5:45] channel cuz why are we even launching, [5:46] Aaron? Why are we launching? What's the [5:48] reason? [5:49] >> We're launching to get users. [5:50] >> Exactly. We're not launching to like get [5:51] a bunch of clout on the internet and [5:53] make people think that we're smart. [5:54] We're trying to actually get people to [5:55] use our thing, and that means having a [5:57] channel that we can keep on repeating [6:00] and find users from. And so, [6:02] uh yeah, I'll show you one example here. [6:03] This one's a perfect example. This is [6:05] like the launch of of a Clicky. So, this [6:08] is Clicky. Clicky is this little blue [6:10] guy that follows your cursor on his user [6:12] computer. And it's basically an AI [6:14] teacher that can teach you anything. [6:16] Honestly, I thought the demo was not [6:18] that good. I remember I showed it to you [6:19] in office hours, and you're like, "Oh, [6:21] you should post this. It's actually kind [6:22] of cool." And [6:24] I I think one day you asked me again [6:26] about it, and uh I ended up saying, [6:29] "Okay, all right, I'll I'll post this." [6:31] Cuz even for me, I've been doing this [6:32] for years, like 10 plus years, and even [6:35] I sometimes I'm like, "Oh, I don't want [6:36] to post it. It's not good enough." But I [6:38] was so wrong. That's so wrong of me for [6:40] you to even think that. And so, I [6:41] recorded this video in the office 5 [6:43] minutes before dinner, and I posted it, [6:45] and after dinner, it's just doing [6:46] numbers. And now this is my company. [6:48] Isn't that That's a crazy thing. It was [6:50] just a random side project that I then [6:51] even just open source cuz I thought no [6:53] one would even care about it. And uh [6:56] here I am working on it 6 months later [6:57] as a real company, right? [6:58] >> Yeah, and so this was nearly 3 million [6:59] views. Uh how how important was this [7:02] video in actually even giving you the [7:03] confidence that you should keep working [7:05] on this? [7:05] >> I do not care about the traffic at all. [7:08] Uh [7:08] the views looks really wonderful, and [7:10] people will say congratulations, but [7:12] what are the actual numbers? And so, I [7:14] actually don't care about this number at [7:15] all. I was like, "How many people [7:16] downloaded it?" And so, after this [7:17] video, I think 7,000 people downloaded [7:19] it. And then I looked at who's like [7:21] retained 4 days later. And if you ask me [7:23] why I'm working on the product today, [7:24] it's not cuz it went viral. It's because [7:26] 7,000 people downloaded it, and about [7:28] 100 people of the 7,000 kept on using it [7:31] every day for 5 days straight. That got [7:33] me so inspired. I reached out to those [7:34] 100 people. I talked to each of them, [7:36] and I was like, "Oh my god, people are [7:38] using this for things I would have never [7:39] imagined." And the users are what [7:41] inspired me to keep going forward. Of [7:43] course, there's so many little things I [7:44] could talk about of things I'm doing [7:46] just subconsciously at this point. Maybe [7:47] you could help me think through it, but [7:49] even just the headline, just like [7:52] just casual. I built this thing. [7:53] >> Yeah. [7:55] >> It's an AI teacher. [7:56] What does it do? [7:58] Why is it important? It's like That's [8:00] how my brain thinks. I've been writing [8:01] for like a decade plus. And so, to me, [8:04] it sounds dumb to call a tweet poetry, [8:06] but like I care about how long things [8:08] are. I care about uh the different the [8:10] different parts of the story that I'm [8:11] telling here. [8:12] >> And I remember this even back to your [8:13] Build Space days when you would send [8:15] your regular marketing emails to the [8:17] community. I would get those, and that's [8:19] how you wrote, too. [8:20] >> Yeah, yeah. [8:21] >> feel like the typical marketing email [8:24] that somebody would expect to get. It [8:26] felt like Farza was writing to me. Even [8:28] when, you know, it it wasn't like you [8:31] directly sitting down and doing the [8:33] thing, it felt like that. [8:34] >> Totally. One thing I do a lot is It [8:36] depends on the platform. This is [8:37] Twitter. I'll just kind of go, I don't [8:39] know, some random person's tweet here. [8:41] Maybe like this one. [8:43] And I'll literally change this text to [8:45] my my tweet's text. Hm. And I'll just do [8:48] this. [8:50] I'll be like, "Did my eyes stop?" [8:51] I'll just ask myself, "Did I myself stop [8:53] on what this is saying?" And if I stop, [8:55] that's like step one for me on testing [8:57] if I myself would kind of check this [8:59] out. And so, that's kind of my number [9:00] one rule for got regardless of Twitter, [9:02] Instagram, YouTube, anything. [9:03] Am I going to stop on this? [9:06] >> Mhm. [9:06] >> Am I even going to read the first [9:07] sentence? Cuz if I don't read the first [9:08] sentence, there's no way I'm I'm going [9:10] into this. Also, like, look at the first [9:12] frame of the video. It's very purposely [9:14] Xcode. Why? Because I know that [9:17] initially curious developers are [9:18] probably going to be interested by this. [9:20] Not like actual video editors, right? [9:22] So, it's purposely Xcode. I purposely [9:24] zoom in right here to this part to show [9:27] cuz the part is so small. The part is [9:29] like this little triangle that kind of [9:31] helps you out. Most people when they're [9:32] scrolling see the first 3 seconds. [9:34] >> Everyone. So, this is click [9:37] >> Zoom. And I go, "Oh, what is that?" You [9:38] know, at least at least while you're [9:39] while you're scrolling. [9:40] >> Yeah, and you're moving it around so [9:41] people can see that there's something [9:42] moving there. [9:43] >> I think that, um, you can go into [9:45] details all day, but the reason I got [9:47] this I know this stuff now is because [9:50] there's hundreds and thousands of videos [9:51] I've made that nobody has cared about. [9:53] Yeah. I think that's the biggest thing. [9:54] Like, yeah, I I don't want people to [9:56] watch this video and think, "Oh, I got [9:57] to make this hook and do this and this." [9:58] >> And so, this was this first one off the [10:00] cuff. This was almost entirely [10:01] unplanned. And so, a lot of this was [10:03] just you and instinct and how you put [10:05] this together based on things you've [10:06] learned. Walk us through another video [10:08] that that maybe you planned out a little [10:10] bit more in advance cuz one thing that's [10:12] surprising to me, you actually write a [10:13] script for all of your videos and I [10:15] think anybody that's seen one of your [10:16] videos would be surprised by that [10:18] because you seem like you're just [10:20] talking off the cuff. [10:22] How do you do that? Like you you're very [10:23] prepared but it doesn't seem formal and [10:26] stuffy and like you're reading. [10:28] >> Totally. [10:28] >> How can you teach somebody else to do [10:30] that? [10:30] >> If I like reading it, then that means [10:32] it'll probably sound good if I say it [10:34] out loud as well. [10:35] >> Mhm. [10:35] >> But not always but sometimes. And so [10:37] this is one example of a script I wrote [10:39] um for a video that I did yesterday. And [10:41] again, it's just super simple like do I [10:43] enjoy reading this? Like would I read [10:45] this if I just someone texted this to [10:47] me? [10:47] >> Yeah, what's going through your head? [10:48] Are you imagining the audience uh [10:50] listening to this or how do you evaluate [10:53] if something is actually good and [10:54] interesting? [10:54] >> For me, it's always I'm the audience. [10:56] Like would I be interested in what what [10:58] the script is saying? And so honestly, [10:59] this script is actually not that strong [11:02] um but I still recorded the video cuz [11:03] you got you got to be launching. In the [11:05] first kind of [11:06] little bit am I getting across what I'm [11:08] doing? Um why is it important? Like you [11:11] can get across so much in just 10 [11:12] seconds. What you're doing, why it's [11:14] important, let's get into it. You know, [11:16] that's what the first 10 seconds of any [11:17] kind of video should be about cuz most [11:19] of these videos are 1-minute long, [11:21] right? It's kind of a movie, you know, a [11:22] movie is an hour and a half, 2 hours and [11:24] the first 20 minutes is just a setup. [11:26] >> Mhm. [11:26] >> It's a 2-hour thing. And so for a video [11:28] that's 1-minute long, you need a setup [11:30] but it should only be like 10, 20%. So I [11:32] write this down and so in this case, I [11:34] said, "Hi, I want to share a new demo of [11:35] our of our new a demo of our new agent [11:37] interface." When talking to our users, [11:39] blah blah blah [11:40] um and then getting into the demo. [11:42] And so [11:43] for me, this is kind of [11:46] this is just kind of how I think through [11:47] it, you know, before I get in front of [11:48] the camera. The worst part is being in [11:49] front of the camera [11:50] and you're like, "Ah, what am I saying?" [11:52] You know, like you have you have nothing [11:54] to ground yourself. [11:55] >> Yeah. [11:55] >> It's good to ground yourself, yeah. [11:56] >> You start with the script and then you [11:58] move on to some of the visuals. How do [11:59] you think about what that first shot [12:01] should be? It's so important, right? [12:02] It's the first thing people see when [12:03] they're scrolling. [12:04] >> I'll show you a quick example of one of [12:06] our videos. So this video actually did [12:08] really good on YouTube. Uh this is like [12:09] a brand new fresh YouTube channel. Like [12:11] we literally made it and then posted [12:13] this video 3 days later. [12:14] >> Yep. [12:14] >> And this is one of our highest [12:15] converting videos, despite being 19,000 [12:17] views. Yeah, this video, I think it did [12:19] like 3 million views on Twitter, [12:20] >> Mhm. [12:21] >> 20,000 on YouTube, and like some crazy [12:23] amount on Instagram as well. The same [12:25] video. So, it kind of did it across We [12:27] posted just everywhere, you know, just [12:28] to see what happens. But yeah, I mean, [12:30] the way I like to do my videos is [12:32] who is the person making this thing? You [12:35] know, like, just show yourself. Show [12:36] your face very clearly. This is you. [12:39] This is your personality. Just be proud [12:40] of who you are, and um [12:42] confidently tell the camera what you're [12:43] working on. Cuz if you think it's cool, [12:46] then I think you'll find at least My [12:48] thought is always at least one other [12:49] person will think this is cool. [12:50] >> Mhm. [12:50] >> And that's who I'm talking to. I'm [12:51] talking to that like one other person in [12:54] that world that may think this is cool. [12:56] And so, yeah, I always start with a very [12:57] clear "Hi, this is what I'm working on." [13:00] >> All you got to do is press and hold at [13:01] 10 plus control, [13:02] >> [music] [13:02] >> start talking, and whatever you're [13:04] saying will stream in as text into [13:06] whatever text box your cursor is on. For [13:08] example, if I'm on Apple Notes, I [13:09] [music] can just say [13:10] >> "Great, that's how it works. Um can you [13:12] give me an example?" [13:13] >> Hey, today I have to ship [music] three [13:14] things. First, skills, second, memory, [13:17] and third, this bug that this one user [13:19] gave about dictation. [13:20] >> Okay, that's pretty cool, but this is [13:22] like every other product I've ever seen. [13:23] There's nothing special about this. [13:25] >> So, that's pretty cool, and honestly, [13:26] many products already do this exact [13:28] thing today. But here's the thing, when [13:30] I'm writing, I often don't really know [13:33] what I even want to say. [music] So, [13:34] check this out. We're shipping [13:35] >> So, then like, I frame the problem for [13:37] myself. [13:38] >> Mhm. [13:38] >> I'm not saying "Everyone has this [13:40] problem where they don't know what to [13:42] write." Like, I'm not involving [13:43] everybody. I'm only talking about [13:45] myself. [13:46] And cuz these are my problems, cuz I'm [13:47] always building stuff for myself. [13:49] >> screen away dictation. [music] It's [13:50] pretty cool. I've never seen anything [13:51] like it before, and I want to show you. [13:53] Let's say I want to write [13:54] >> Again, that's just me being myself, [13:56] which is like, "Hey, I want to show [13:57] you." [13:57] >> You come across very natural here. Um [14:00] was that easy for you to do that in the [14:01] beginning? [14:02] >> First of all, this whole video is [14:03] scripted. Every every line that I'm [14:05] saying, I wrote like 3 days before. [14:07] >> Mhm. [14:07] >> And I said it and I I I I will say it in [14:09] my head like over and over and over [14:10] again, in the shower, uh [14:13] going to work, and I'm always like, "How [14:14] should I say that line? How should And [14:15] by the time I'm in front of the camera, [14:17] I've already visualized it like 10 [14:18] times. [14:19] And it's so funny, after this video we [14:21] were finished it, I told Vikrant, uh, [14:23] the person [14:24] uh, who I make videos with, and I'm [14:26] like, this is the worst video we've ever [14:27] done. Uh, it's not even worth editing [14:29] it. And he's like, no, bro, be quiet, [14:30] shut up. Uh, I'm going to make it and [14:32] then we're going to post it. And so [14:33] again, it's good to have the team also [14:34] be really like, hey, we made this, like, [14:37] we're going to put it out. It doesn't [14:38] matter how cringe we think it is, like, [14:40] we'll let the audience decide. This [14:41] video did 3 million views on Twitter. [14:43] And people went crazy over it. And it's [14:45] so funny, I thought it was horrible. I [14:48] just thought like, this is so this is so [14:49] bad. So this is one of the situations [14:51] where I was wrong. And I think founders, [14:53] I mean, me especially, think something [14:55] we made is bad, or I made is bad, and [14:57] then I don't want to put it out cuz it's [14:58] going to make I don't like how it sounds [15:00] or feels, I'm constantly forcing myself [15:02] to put it out. It's not natural for me [15:03] to click post. Um, [15:05] but uh, sometimes it is. [15:07] Sometimes it's not. Yeah. And so this is [15:08] one example where I was like so [15:10] surprised [15:11] uh, that people even cared to begin [15:12] with, cuz I thought it was just a pretty [15:14] simple thing. [15:15] >> It's interesting, um, I I've heard you [15:16] talk about this almost like demo-driven [15:18] development where everything that you're [15:20] working on, you're almost working [15:21] backwards from a demo that you want at [15:23] the end of it. And that's kind of the [15:25] proof that the thing is ready to go, it [15:26] lets you test it out, and then you get [15:28] to the end of it and you're like, let's [15:29] just publish the demo. [15:31] >> The most important thing in my company [15:32] that I try to always and every company [15:34] I've ever run is, you know that feeling [15:36] when you're working with someone, often [15:37] on the team or a co-founder, and you're [15:38] like, yo, check this out. That, you have [15:41] to like be so aware of that, cuz that is [15:43] the thing that you could probably talk [15:44] about online. That check this out is [15:46] exactly like that raw check this out, is [15:48] exactly how you should be kind of [15:50] presenting yourself on the internet. In [15:51] this case, you know, sometimes I go a [15:53] little bit more fancy if I want to, but [15:55] sometimes I'm also just a guy in a a [15:57] simple video, and in the YC office. And [15:59] then sometimes, maybe once a quarter, [16:01] we'll do a polished launch, where it's [16:03] like, this is like something I'm really [16:05] proud of, I want to make sure it's [16:06] talked about properly. But generally, [16:08] we're launching like [16:10] demos every single day almost. [16:12] >> It's interesting, now that everybody [16:13] can, um, put out videos and and do [16:15] launches every single day, how do you [16:18] stand out in that world? It seems like [16:20] um [16:20] people that are spending lots of money [16:22] and trying to do fancy produced videos, [16:24] that's an effort to try to get more [16:26] attention and stand out and and seem, [16:29] you know, polished and and further along [16:31] than you are. Um but you take the [16:33] opposite approach. And so in a world [16:35] where there's just so much content and [16:36] so many videos that are out there every [16:38] single day, how do you think about how [16:40] you try to stand out and how others can, [16:42] too? [16:43] >> I never think about like that. I always [16:45] just think about like [16:47] I'm I I want to talk about something. I [16:49] think it's interesting. And there's got [16:50] to be one of the person out there that's [16:51] that thinks it's interesting. [16:52] >> Mhm. [16:52] >> And when I click post, in my brain, I'm [16:55] always I'm always like if this gets zero [16:57] attention, zero views, I'm still happy I [16:59] posted it. You know, because I'm glad I [17:00] got something out there. I think that [17:02] should be the mentality everybody has [17:04] working on their company [17:05] launching 2026. Like you can't be [17:08] thinking about it. By the time I hit [17:09] post, I've already forgotten it. Um I'm [17:12] looking at it maybe a week later. [17:13] >> Yeah. [17:14] >> Yeah. [17:14] >> How do you do that? I feel like most [17:15] people when they post something they're, [17:17] you know, incessantly just refreshing [17:19] and trying to see are people liking it, [17:20] who liked it. Um but you launch and you [17:23] walk away. [17:24] >> Yeah. [17:24] >> you get to that point? [17:25] >> I mean, you just put so much garbage out [17:27] there. I put so much garbage out in the [17:28] world at this point. Not like bad stuff, [17:30] but just like stuff that I wasn't like [17:31] this is I didn't think that I didn't [17:32] think was good. [17:33] >> Yeah. [17:33] >> And [17:34] people forget the next day. I can make [17:36] something right now and post it and [17:37] people are going to forget the next day. [17:39] That's just how [17:40] attention in 2026 kind of works, you [17:42] know? [17:42] >> And so this is like a perfect example. [17:45] This is 2021. Mhm. It's a year after YC. [17:48] I'm like a I'm a solo founder. I've been [17:49] a solo founder at ZipSchool and a solo [17:50] founder at Buildspace, solo founder now. [17:52] It's not like I'm that Yes, I know how [17:54] to put a camera in front of me and edit [17:55] it, but I'm not a fantastic editor. And [17:57] so this is back when I was working on [17:59] ZipSchool, which is an online elementary [18:00] school, and I wanted to show off the new [18:02] classes that we were launching for our [18:04] users. And this is the launch video. [18:06] >> Hey there, ZipSchoolers. Faraz from [18:07] ZipSchool here again to show off four [18:09] new classes that we just added yesterday [18:11] to ZipSchool that we are very excited [18:13] about. The first one is called Mars [18:15] Perseverance Rover Landing aimed at kids [18:17] ages 7 to 12. And this class covers the [18:19] Mars landing that actually Honestly, it [18:21] is not I'm already checked out watching [18:23] this video. And [18:24] >> What would you do differently? What are [18:26] the mistakes that you're making in this [18:27] video? [18:27] >> I think that's the important part. It's [18:28] not about that. It's about the fact that [18:30] I posted it and then it got 23 likes, [18:33] which by the way is more than zero. It [18:36] got more than zero views. It got more [18:37] than zero comments. It got one comment [18:40] from a guy. [18:41] >> You got a bookmark. [18:41] >> I got a bookmark. So like I launched. [18:44] This to me is a launch. And I know that [18:45] sounds dumb to maybe some, but to me [18:48] this is exactly what launching means. We [18:50] did something new at the company. We [18:52] talk about it to the into the world and [18:54] it doesn't matter if we got 23 views. If [18:57] there's a chance that we get more than [18:58] one view, we should talk about it. And [19:00] so this is a super small thing. This is [19:02] like an everyday thing in my company [19:03] where we have cuz we have new classes [19:05] dropping everyday almost. And so I just [19:07] make it I'd be like, yeah, let's talk [19:08] about it. Maybe one day [19:11] I'll talk about it and one of the [19:12] classes will be really interesting and [19:14] get picked up and go super viral, which [19:16] never happened. But the fact is I made [19:18] the videos. And over time I I got so [19:20] much better at just knowing like how [19:22] should I talk about products? Cuz think [19:23] about it like I'm good at making videos [19:25] in general and writing stories, but [19:28] talking about products is hard. How do [19:30] you come off without sounding schilly? [19:31] How do you come off without sounding [19:32] fully yourself? And it's hard. This is [19:34] funny. This is a launch. This is the [19:36] first day I worked at my company, [19:37] December 1st, 2020 and I shipped a [19:39] landing page. Aaron, there's no copy. [19:42] There's just a link and there's some [19:45] weird video I made in Canva and I posted [19:47] it. [19:48] And that's a launch. To me I launched a [19:50] landing page. Right? In my eyes. And by [19:52] the way, in this instance I had like [19:53] maybe a 80 followers or something. So no [19:55] one was watching what I was doing. And [19:57] this is just continued. What is this? [19:58] Nice four years. [20:00] >> Uh I'm pretty excited for the next four [20:01] years or [20:04] the USA [20:05] for ZipSchool [20:06] for the world. [20:08] It's going to be great. [20:10] >> And so again, I've just been posting [20:11] these for just years. And so over time [20:13] you you build up You know, it's like [20:15] it's like basketball. You just keep on [20:18] keep on going like, "Oh man, this is [20:19] getting closer." And you internally are [20:21] smart are human being, you're smart [20:23] >> Yeah. you'll slowly get better. But if [20:24] you don't take a thousand shots, you're [20:26] not going to make the that that the the [20:28] 5,000th shot, you know? [20:30] >> Clearly, your advice is just put it out [20:33] there. Don't worry about it. Just put it [20:35] out there. You need to get the reps. [20:36] What are the things that you've learned [20:38] in the five years since then? [20:40] >> So much. [20:40] >> The tips that you could give people now [20:42] in order to improve their videos and [20:44] start from a better starting point than [20:46] you did when you were first doing this. [20:48] >> Totally. I want to be really clear. All [20:49] my videos are like look very produced, [20:52] uh but that's because it comes very [20:53] naturally to me. To make a produced [20:54] video for me will take like an hour. Um [20:56] but for other people may take longer. [20:58] So, while I'm showing this, I kind of [20:59] want to be clear like you don't have to [21:01] produce your videos. It could just be an [21:02] iPhone video and it'll still do really [21:03] well. So, this is one of those videos [21:05] where I just knew even when I when I [21:07] when I wrote the script that it was [21:08] going to go big. [21:09] >> Mhm. [21:09] >> Uh it has now 21 million views on [21:11] Instagram. And believe it or not, this [21:12] one video is still driving a lot of our [21:14] traffic. So, that's how these algorithms [21:17] work like they'll keep serving the [21:18] content to people uh to more people that [21:20] want to see it. Also, I've made this [21:22] video an ad now. So, I'm running [21:23] Facebook ads now on a video. So, yeah, [21:25] there's a lot of benefit to making [21:26] videos for your company. [21:27] >> Last week, I went super viral for an [21:29] idea that I built that I thought was [21:31] actually kind of terrible. And this [21:32] week, I'm somehow in Y Combinator? Let [21:35] me show you what [21:35] >> Okay, so just that hook. [21:36] >> Yeah. [21:37] >> Like what what did I say? Um last week I [21:39] went super viral for something I built [21:40] that I thought was kind of ridiculous or [21:42] whatever, bad. Um cool, like that's [21:45] interesting. And now he's in this thing [21:46] called Y Combinator. I've heard of Y [21:48] Combinator. So, I was kind of targeting [21:49] that niche. And then a very simple, let [21:51] me show you what I'm working on. I had [21:52] it's My formula is always the same. [21:55] >> Yeah. [21:55] >> This is who I am. This is what's going [21:57] on. This is what I'm working on. Let me [21:59] show you. [21:59] >> Show you what I'm building. It's called [22:01] Clicky. It's a little AI buddy that [22:02] lives next to your computer cursor and [22:04] teaches you anything in real time for [22:06] free. Let's say I'm learning how to [22:07] color grade in DaVinci Resolve. Instead [22:09] of going to a YouTube video, I just got [22:11] to ask out loud and Clicky will [22:12] literally tell me exactly how to get [22:14] there and even point at it for me. [22:15] >> So, again, a little pretty produced, but [22:17] again, this this took us like four to [22:18] six hours to make because we just had [22:20] done this. But again, even if this was a [22:22] simple video where I'm like, "Hi, this [22:24] is what I'm working on. Check this out." [22:26] And then I turn the camera to literally [22:28] show my computer. I think the effect may [22:30] not have been as big, but again, you [22:32] just start and you make like 10 of those [22:34] then. And you get really good at talking [22:35] about stuff. But again, like that video, [22:38] super clear hook, a very clear like why [22:41] you should why you should care right now [22:42] while you're on your scroll. And so a [22:44] big thing that I always do when I watch [22:45] videos on the internet, regardless if [22:47] they're mine or not, is I ask myself, am [22:49] I bouncing? When am I bouncing? And so [22:52] if I go to Instagram Reels anytime, and [22:53] this is what I do a lot actually. I [22:55] watch a lot of content, not for [22:57] entertainment, just to kind of study. [22:59] And so I can go to Instagram Reels. [23:01] >> page to block my long-distance [23:02] girlfriend and post my weight loss [23:04] journey. [23:05] >> First of all, this is a random video we [23:06] just stumbled upon, right? There's so [23:08] much to learn. [23:09] Dark background, light on his face, kind [23:12] of a cool little caption at the bottom [23:14] where it started 2 months ago. You kind [23:16] of want to know 2 months ago what? [23:18] You're just curious. [23:19] >> Right, what happened? [23:19] >> And so I'll mentally take that as a note [23:21] and be like, "Oh cool, like maybe we can [23:23] make a video that uses this as like a [23:25] template. Maybe it's like 2 months ago I [23:27] was working on something and I didn't [23:28] really know what it was and then tell [23:30] the story." And so again, I'll just keep [23:32] doing this. [23:34] >> Okay, so this song just hit 2 million [23:36] streams. [23:36] >> You had a spell when you [23:38] >> You wanted to know where I've been. [23:40] >> It's all Welcome back to features we [23:42] should have released 6 months ago, but [23:44] we're un-releasing them now. [23:45] >> This is cool. This is what startups [23:46] should be doing. Yeah, so I'm seeing a [23:48] lot of this now, but mostly like younger [23:49] founders, which is cool. [23:50] >> There's almost something [23:52] like self-deprecating about it. You [23:55] know, yours is like, "Ah, I had this [23:57] thing that I thought was terrible." And [23:58] they're like, "Ah, this is the stuff we [24:00] should have launched 6 months ago." And [24:01] this guy's like, "I broke up with my [24:03] ex-girlfriend." [24:04] >> Everyone wants a story. And so how can a [24:06] demo just be a mini story or what you're [24:08] saying be a mini story? [24:09] >> Yeah, and you I've heard you talk about [24:10] this where it it's not actually you in [24:12] the video, it's the character Farza. And [24:15] and it's almost like all these people [24:16] they're a character that the audience [24:18] can identify with and [24:20] ultimately root for potentially. [24:22] >> Yeah, yeah. I mean, I think that like [24:24] when a founder talks to the camera, like [24:26] you come off as a certain type of way. [24:28] And so, um totally, sometimes I'm like, [24:30] yeah, like I am not playing a character, [24:31] but like I'm a character to the people. [24:33] Like when someone watches me on their [24:35] little TikTok feed or their Twitter [24:36] feed, like I'm just like random [24:37] character in their eyes. I'm not like a [24:39] real I'm a person, but I'm kind of a [24:40] character. I'll show you guys this, too, [24:42] which is just something we've been doing [24:44] that our users have been loving. Uh it's [24:48] not going viral, but our users have been [24:50] loving it. And it's just these little [24:51] Actually, some of them have been going [24:52] viral, which is pretty cool. This is [24:53] everything we built this week. [24:56] >> So, ClickUp had been uh had some [24:58] problems about like the integration and [24:59] orchestration of the agents. Now, [25:01] ClickUp has a new brain, and it will [25:02] feel way more smooth to spawn the [25:04] agents, to control them, to pause them, [25:06] and you're going to see a lot of [25:08] improvements overall. A lot of people [25:10] complaining about like ClickUp is keep [25:12] following uh them, so I made a fix for [25:15] them. [25:15] >> So, there's no there's no hook. There is [25:17] no attention. There's no story. It's [25:19] just we're a group of people, and we [25:21] want to tell you what we're doing. This [25:23] almost has a million views on Instagram. [25:25] >> Why do you think these are so popular? [25:27] >> I think that this is the most important [25:29] thing I think about like companies [25:31] making content. What we are doing on a [25:33] day-to-day basis at our companies is so [25:37] far-fetched from what most people think [25:39] is actually happening at a company. And [25:41] uh to just give them an inside look [25:44] of what's actually going on in your [25:45] process is so cool, and it gets them [25:48] excited. It gets them interested in what [25:49] you're doing. What I'm so against and so [25:52] anti is talking about my life as a [25:56] founder on the internet. You will never [25:58] catch me making those videos about this [26:00] is a this is my day as a YC founder. [26:02] Never. Uh like I'm not trying to like [26:04] tell you about my life and tell you [26:06] about like my my my that stuff. I'm [26:07] trying to show you like what I'm doing [26:09] and why [26:10] uh in new ways. And I think that um most [26:12] companies just never show their process. [26:14] For most companies, [26:15] uh big companies, Apple, Meta, you just [26:17] see Mark Zuckerberg on the stage, and [26:19] he's showing you his Meta glasses. But a [26:21] lot of people, especially early adopters [26:23] of of products that uh you know, that [26:25] I'm sure many YC founders are working [26:26] on, uh are interested in just like, "Who [26:29] are these random group of four guys and [26:30] girls that are making this AI product [26:32] and how are they doing it? And just [26:34] telling we make these videos in 10 [26:36] minutes and it's just hey, just get up [26:39] what are you doing today? And yeah, it's [26:42] interesting. Um it's funny when we [26:43] started making these we just made them [26:45] just to stay accountable. Uh cuz it was [26:47] uh it was good to kind of tell the [26:49] people what we kind of stand up but in [26:51] front of the world and so uh that was [26:53] kind of the idea and I mean now we now [26:55] we do these every two weeks and it's [26:56] been really fun. [26:56] >> Yeah, that's really cool. Um I I think [26:58] another thing that comes across here is [26:59] not even just the words that you're [27:01] saying but like the joy that is in the [27:03] videos and I think um people see that in [27:06] your personality and I think they see [27:07] that in the music that you choose and [27:10] everybody's kind of like clapping and [27:12] supporting each other as they're giving [27:13] their update and it's just like positive [27:16] vibes are coming through the video and [27:18] how do you capture that? [27:19] >> Uh see this is so weird because this is [27:21] where you just have to be yourself. I [27:22] keep saying that. People always say how [27:24] do I start making videos about my [27:26] company or videos about what I'm working [27:27] on or awesome presents. It's always the [27:30] same feedback. Make take out your [27:32] iPhone, click record, go go go uh [27:36] selfie mode and talk about what you're [27:38] doing for 30 seconds and post it to your [27:39] Instagram. Post it to your TikTok and [27:41] then post it to Reddit and Instagram or [27:43] Reddit and YouTube or whatever. And you [27:45] know how many people take me up on that? [27:47] Almost like maybe 1% 2% of the time [27:49] people will actually do it. Why though? [27:51] If I told you right now I mean maybe [27:52] you're different I don't know I don't [27:53] you're probably different you're you're [27:55] Aaron Epstein uh but if you if I tell [27:56] you to point your phone and record a [27:58] video uh how does that make you feel? [28:00] And then post it to your personal [28:01] Instagram. What do you do? [28:02] >> you get nervous about it like what am I [28:03] going to say? How am I going to look? [28:05] What if people think it's stupid? And [28:08] all these thoughts I think go through [28:09] people's heads and they're they're [28:10] afraid of [28:11] >> high school friend follows me. He [28:13] doesn't care about this what I'm working [28:14] on like nobody cares. But see that's the [28:17] thing like there's so much value to [28:20] you got to be cringe for a little while [28:22] actually to learn how to not be cringe. [28:24] And so a lot of people will like post [28:26] their first video and be like this is [28:27] this is horrible I'm not going to post [28:29] this. But then yeah I think the [28:30] important part is to continue doing it. [28:31] And so again [28:33] every day as a founder, founders [28:34] watching on YouTube at YC, you're [28:37] spending 8 hours on your laptop making [28:39] magic, smiling, seeing things that you [28:41] find are interesting. If you just spend [28:44] even 10 minutes to step back and ask [28:45] yourself, "What's something I did today [28:47] that made me smile or was interesting or [28:49] felt new?" And just talk about it, and [28:51] then show it, and then click post, and [28:53] then close it, back to work. That's all [28:56] you have to do. That's what it means to [28:58] be launching, I think, in 2026. Just to [28:59] show your work, cuz that is what [29:00] launching is, right? When you launch on [29:02] TechCrunch or New York Times, like, [29:03] you're showing your work. But in 2026, [29:06] you show your you can show your work [29:07] every day. And you can actually get an [29:09] audience just by showing your work every [29:11] day. [29:12] >> of people think, "Well, I'm not [29:13] targeting Gen Z or whoever [29:16] stereotypically TikTok users, Instagram [29:18] users. Therefore, I should not use that [29:20] as a channel to try to acquire my [29:23] business customers or I sell to [29:24] procurement or whatever." But it sounds [29:26] like your argument is like, "That's not [29:28] true. There's all kinds of people using [29:30] these platforms. They're real humans [29:32] that happen to work at companies where [29:33] they could be your customer [29:35] >> Oh, yeah. [29:35] >> and would be interested in seeing your [29:36] products, right? [29:37] >> Yes. Everybody in the world is on [29:41] Well, not everybody in the world, but a [29:42] large majority of the world [29:44] is browsing through these platforms. [29:45] >> Billions. [29:46] >> Yes. We found that design agencies [29:48] really loved our product. Like, three of [29:51] them that found us naturally. So, the [29:52] question became, how do we get more [29:54] design agencies? Now, well, well, I'm [29:56] curious, what what what what what would [29:57] you do if you take yourself back like to [30:00] founder mode Aaron? [30:01] >> Oh, I don't know. Probably the default [30:02] for a lot of people is like, go look on [30:04] LinkedIn and, you know, try to find [30:05] people that list the place they work as [30:07] a design agency or go to communities [30:10] that they hang out in or like [30:11] specialized design [30:13] >> Totally. [30:13] >> websites or something. [30:14] >> And those are all actually pretty good. [30:16] You should do those things. But at the [30:18] same time, do you think a design agency [30:20] founder is on TikTok or Instagram or [30:21] YouTube? Do you think they're on these [30:23] platforms? The answer is probably yes. [30:25] It's probably a lot of them. And so, we [30:26] started making TikToks, I remember. I [30:28] don't have the the account pulled up [30:29] here that was just targeting agencies uh [30:31] design agencies. I would literally say, [30:33] "Hey, are you a design agency? Or or or [30:36] hey, are you a the founder of a design [30:37] agency? Stop right there. Let me show [30:39] you something really cool. And then [30:41] that's how we found like 50 design [30:42] agency leads [30:44] just through TikTok. And these videos [30:45] didn't go viral. They each on average [30:47] had like 4,000 views. [30:48] >> But it was about 4,000 targeted views. [30:50] >> 4,000 targeted views. So, the goal of my [30:52] goal is never to go viral. It's just I [30:55] want to talk about what I'm working on [30:57] and I want to make sure it's listened to [30:58] by an audience that would be interested [31:00] in actually interested in what I'm [31:02] doing. [31:02] >> How do you think about the different [31:03] channels? [31:04] >> Every channel is different. I'll say off [31:06] the bat, Twitter is our worst is in [31:08] historically I've I've been making stuff [31:10] on Twitter for like 10 years. [31:11] >> Mhm. [31:11] >> The worst conversion of any channel. [31:13] Like this video, the original clicky [31:15] video, got 7,000 downloads on 3 million [31:19] views. [31:19] >> Mhm. [31:19] >> If you do the math there, it's like less [31:21] than 1% or some crazy low number. [31:22] >> Much less, yeah. [31:23] >> It's very low. Screen aware dictation, [31:25] this video, right? 2.9 million views on [31:28] Twitter. But the conversion rate onto [31:30] those people actually making accounts [31:31] was horrendously low. This YouTube video [31:35] has 20,000 views. [31:37] But do you know what the conversion rate [31:38] on this video is? It's like 20% to them [31:40] actually downloading the app. [31:41] >> Wow. [31:41] >> So, the chances of you watching this [31:42] video and downloading my app is 20%. [31:44] >> Do you think that's something specific [31:46] to the channel or is it something [31:47] >> Yeah, I know. Every channel is [31:48] different. Think about it. I'm on [31:50] Twitter. I'm just like doom scrolling [31:52] basically. Not really doom scrolling. [31:53] I'm just scrolling through random stuff [31:54] all the time. And so, for YouTube, I [31:56] don't know I it was so surprising. And [31:58] so, historically actually YouTube is for [32:00] most companies the best performing [32:02] channel in terms of conversion. [32:03] >> But [32:03] >> our numbers are very low. 872 [32:05] subscribers, 20k views. Like I'm not Mr. [32:08] Beast, you know? But again, if you do [32:10] the math, this leads to a very real [32:12] number of signups. And this video will [32:14] keep on performing. That's the most [32:16] important part. [32:17] >> Yep. [32:17] >> This YouTube is a library that will last [32:19] for decades. Twitter, you got your [32:21] moment once and you're gone. [32:22] >> Mhm. [32:22] >> Instagram is another one that's so [32:24] important. [32:25] All of our traffic I think 70% of our [32:27] traffic is Instagram. This is where you [32:29] got to be smart. You're like, I'm not [32:30] trying to go viral. I'm trying to make [32:33] tell people what I'm doing. But I'm [32:34] again like down the funnel, how many [32:36] people actually converted? [32:37] >> Mhm. [32:38] >> It's worth checking out. And so, that's [32:40] what [32:40] I think people should always make their [32:42] video and then start make the first [32:43] couple of videos, but then start to ask [32:45] themselves, "Hmm, maybe I should make my [32:47] For example, our link in bio is {slash} [32:49] IG. Mhm. Why? Because then I can track [32:51] like wait a second, what was the [32:53] conversion rate of people who hit [32:54] {slash} IG and downloaded my app?" Or if [32:57] they landed on their phone, [32:58] sent it to We have a little button that [33:00] lets them send it to them themselves [33:02] over email. How many of them clicked [33:03] that? What was the conversion rate? And [33:04] so, by posting on all these channels, [33:07] you start to see not only where I'm [33:09] performing views-wise, but actually just [33:11] like who's Where is my best channel [33:14] right now in terms of the audience that [33:15] I'm looking for? And again, for us, it [33:17] is by far Instagram, [33:19] and it's it's not even close. I think [33:21] you're basically using the same content [33:23] for every channel. Yes, I I do make [33:26] videos for specific platforms. So, most [33:28] of our videos we make for Instagram, but [33:31] then we post to Twitter. Whereas some [33:32] videos we make for Twitter and then post [33:34] to Instagram. [33:34] >> What is the difference between one that [33:36] targets one versus the other? [33:37] >> So simple. I mean, Twitter generally [33:39] doesn't like vertical videos. It likes [33:41] 16:9. So then this For example, we have [33:43] this drawing video. So, we built a [33:44] system that can draw directly on your [33:46] screen. It can draw polygons, arrows. [33:49] This is super optimized for Twitter. Cuz [33:51] just look, let's let's just watch it [33:52] muted and imagine I'm scrolling my feed. [33:54] Ready? [33:56] Okay, guy in front of blackboard, AI [33:58] draw screen, I'll watch. It's like it's [34:00] that quick. It's like AI draw screen, [34:03] blackboard, he's drawing things, Cloudo [34:05] Fest demo, click. You know, like my [34:07] brain thinks like that. What are the [34:09] quick Cuz nobody's reading this entire [34:11] this entire damn thing. And so, this is [34:13] optimized for Twitter. This video is [34:15] optimized for Twitter. But, what was [34:17] surprising is we posted the exact same [34:19] thing to my Instagram, and it performed [34:24] even better. And you would think that [34:26] like this type of video wouldn't perform [34:28] that good, but it has like millions of [34:29] views. And it's a vertical video cuz on [34:31] their phone, it looks even worse. It [34:33] looks like this. [34:33] >> Well, one thing that I maybe a couple [34:35] things that I know about you are you [34:37] have a very high bar for yourself and [34:39] the things that you put out there. and [34:41] while at the same time almost ironically [34:43] you're very comfortable putting a lot of [34:45] things out there. [34:46] >> Yeah. [34:46] >> Um and so you [34:48] I hate to use the word taste, but like [34:50] you've been doing this so long that you [34:51] have a high bar and a good sense of of [34:54] what you would be excited to watch and [34:57] you just naturally have a very short [35:00] attention span. You get bored and you [35:01] like want the next thing and if you're [35:03] not getting that then you move on and I [35:05] think that serves you very well with the [35:07] videos that you're creating here. [35:09] >> Yeah, either I have good taste or I just [35:11] put out so much stuff and some of them [35:13] just hit. [35:13] >> That's how you get it I think is you you [35:15] put it out you put out a lot of things [35:17] and you continue to improve and that [35:19] that levels up your work. [35:20] >> Totally. I think I have more the latter [35:21] where I just put out so much stuff and [35:22] some of the some of it works and [35:24] again we're looking at the stuff that's [35:25] working, but so much hasn't worked. [35:27] >> Right. [35:27] >> Like my old Zip School videos like [35:28] nobody cared. [35:29] >> Yeah. [35:29] >> But this is what this is a perfect [35:30] example. This was a super simple video. [35:34] Basically, just a quick story here. [35:36] I found out that with GPT real time you [35:38] could generate shell commands. I can [35:40] control your computer in real time. And [35:41] so we can watch it real quick. All [35:42] right, check this out. No hands. Ready? [35:45] Hey Cookie, can you please open up [35:46] Spotify and play ACDC for me? Maybe uh [35:49] Back in Black. I love that song. [35:54] Rock on. It's playing Back in Black now. [35:56] Amazing amazing. Cookie Hey Cookie, it's [35:57] a little bit loud. Could you turn it [35:59] down in Spotify 50%? It's so loud. [36:01] >> [music] [36:02] >> Got it. Lowering Spotify to 50% so it's [36:05] less intense. Okay, thank you. Hey [36:07] Cookie, can you open up my So now I'm [36:08] going to keep going. And this is [36:09] actually a fun one. Stripe dashboard for [36:11] me in my browser. Let's see if we made [36:12] any money today. [36:16] Opening your Stripe dashboard now. [36:19] Oh man, $140 in a day is not And so to [36:22] me that was just one little moment in [36:25] the day where I was like, oh my god this [36:27] AI model could do this. [36:29] I didn't [36:29] I didn't check if anybody did it before. [36:31] I was just like, I'm going to make a [36:32] video. Turns out I did something very [36:34] novel. And um again the examples I used [36:37] playing a song on Spotify [36:39] uh and lowering the and lowering the [36:40] volume that's exactly the first use case [36:42] that I had as a user myself. [36:43] >> Yeah. [36:44] >> And so again, always building for like [36:46] the version of myself that's kind of [36:47] watching out there. Cuz if I think [36:48] that's cool, I think someone else would, [36:49] too. And hey, look. [36:51] 3.8 million views on Twitter. [36:53] Really good. But again, then Instagram, [36:56] 4.6 million. You know, people always say [36:58] like you got to be super like your brain [37:00] has to be rotten when you're scrolling [37:02] Instagram, and you got to have a lot of [37:03] quick edits and blah blah blah. Like [37:05] there's not a single cut in this video. [37:06] >> Yeah, there's no cuts in the video, but [37:09] there are beats in the video. [37:10] >> Yeah. [37:11] >> That you do a very good job with. You [37:12] know, [37:13] we see you first. You say, "Hey, no [37:15] hands." So right off the bat, I'm like, [37:17] "Oh, no hands. He's just talking to this [37:19] thing." [37:19] >> Yeah. [37:20] >> Which is the hook that I think gets [37:21] people to watch. [37:22] >> That's all purposeful. [37:23] >> Yeah, then you play the song. Then you [37:25] lower the music so that you can continue [37:27] to talk over it, and it responds to that [37:28] as well. Then you pull up the stripe [37:30] stripe dashboard, and somebody goes, "I [37:32] wonder how much money they made today." [37:33] >> There's always there's Yeah, I'm always [37:35] just trying to keep the story moving. [37:36] I'm not trying to keep your attention. [37:37] That's actually the wrong way to look at [37:39] it. I'm always trying to I'm always [37:40] like, "All right, where's this guy [37:41] going?" Come on. You know, you're like, [37:43] "Come on, keep going." When you're [37:44] someone talking, like Let's get to the [37:46] end of the story. Okay, you went to the [37:47] movies, and I said you just keep doing [37:49] that endlessly. And so [37:51] um I I love that you said the word [37:53] beats. Some of my videos are purposely [37:55] made [37:56] to just kind of like appeal to like [37:58] the audience that just [38:00] has no attention span. [38:01] >> Mhm. [38:02] >> You know, for example, like this one. It [38:03] has been 9 weeks since I built and [38:05] launched my idea to the world. [38:07] It's an AI that can draw directly [38:08] [music] on your screen and guide you [38:09] through any So that's one example of [38:11] like a quick. [38:12] >> Yeah, yeah, yeah. [38:13] >> I'll show you this example here, which [38:14] is um [38:15] talking about the beats just for once. [38:17] >> Two weeks ago, I was in my apartment [38:18] trying to figure out a cool idea to [38:19] [music] work on. Two weeks later, I [38:21] built this thing that's now being used [38:22] by thousands of people every single day. [38:24] And the crazy part is [38:25] >> So again, I literally think about it [38:26] like like like this. [38:28] >> Mhm. [38:30] >> Am I here? [38:31] >> Yeah. [38:31] >> Am I still watching? I think as you make [38:32] these videos again and again, um it [38:34] becomes a lot more obvious to you as you [38:36] do it. Uh we have this one guy at our [38:38] office right now. He is 19, and he has [38:41] this built into his head. Uh the same [38:42] the same beat system. And it's It's like [38:44] most people who work on videos a lot [38:46] have this little beat system in their [38:47] head. [38:48] >> I think another thing that's unique is [38:49] you actually um invest a lot in your [38:51] early team, which is small, but a large [38:54] portion of it is actually built to help [38:56] you create and produce and get these [38:58] videos out the door on a regular basis. [39:00] >> Yeah, I mean I always say like I'm okay [39:03] at videos. There are people out there [39:05] who who are way better at me. [39:06] >> Mhm. [39:07] >> My first two hires were I knew from the [39:09] beginning, even before I started my [39:10] company, it's going to be number one, a [39:13] filmmaker, and number two, an engineer. [39:16] And this number one is film is filmmaker [39:17] for me. So that is the first one. [39:20] Or whenever whatever order they can come [39:22] in. [39:22] >> Why was that so important to you? [39:23] >> I don't want to put a ton of attention [39:25] every day on making videos. I mean any [39:26] founder watching, any founder in YC, [39:28] like nobody wants to do that every day. [39:30] At at some at some level you do have to [39:32] in the beginning in the early days. And [39:33] yeah, I mean look at my Clickey videos, [39:35] like I'm forcing myself to just make [39:36] videos and make stuff every day. At one [39:38] point, you're like, well, this would be [39:40] really great if someone's full-time job [39:42] at my company was to kind of observe [39:44] what we were doing and help us launch [39:46] and constantly make it their [39:48] their mission to make us look good when [39:49] we launch as well and make us come off [39:51] properly. [39:52] And so that's why a filmmaker was one of [39:54] the first hires I I made at this [39:56] company. And at my last company, our [39:57] film team was two people. We were only [39:59] like a seven-person team. [40:00] >> Yeah. [40:00] >> So even at this company, at one point at [40:02] Clickey, we had three people making [40:04] videos and one engineer. Um cuz it's so [40:07] funny now, it's so easy to pump out more [40:10] code and more software, but it's harder [40:12] than ever to actually talk about what [40:15] you are doing in a coherent way in front [40:17] of an audience. [40:17] >> Mhm. [40:18] >> Cuz now Aaron, you know, someone can [40:20] clone my product in like two minutes at [40:22] this point. But again, what they what's [40:24] really hard to clone is the audience and [40:26] the community that I built around what [40:27] I'm doing and the energy around it. If [40:29] you got an audience around your company, [40:31] I'll say 100,000 people that was really [40:33] into what you were doing, maybe [40:34] enterprise AI, you know, uh maybe [40:37] something around like photography or [40:38] whatever, whatever your niche is, [40:39] >> Mhm. [40:40] >> that'd be useful. Cuz cuz that's why you [40:42] know, I can make a video now about a new [40:43] feature and like thousands of people [40:45] flood my app. [40:46] That's awesome. That's like every [40:47] founder's dream, right? [40:49] >> Tell us more about the gear that you're [40:50] using today to to produce these videos. [40:52] >> Yeah, so for demos on my computer, which [40:54] is just like what I'm hacking on that [40:55] day, I always use this thing called [40:57] Screen Studio. I think it's like made by [40:59] some indie hacker. It's just really [41:00] good. It's [41:02] record your screen, record your local [41:03] audio, makes your cursor look good, and [41:05] I think it costs like 30 bucks. That's [41:07] like one level. Then the next level is [41:09] like, well, videos that more include me [41:11] in a more kind of like [41:13] kind of native way, and for that I just [41:15] pull up my iPhone. Literally just a [41:16] selfie camera and just record myself [41:18] talking, post, or edit on my phone even. [41:20] Um I use on my phone it's Instagram's [41:23] editing tool, it's called Edits, a lot. [41:25] So when I I I do edit on my phone a good [41:27] amount. So that's the iPhone. [41:29] Nothing else required. But then there's [41:30] like the next level where it's like you [41:32] feel like you want to do something a [41:34] little bit more kind of like thought [41:36] through, cinematic, and that's where I [41:38] use my Sony FX3. Sony FX3 or Sony FX3? [41:42] The cheaper one. Along with a DJI mic. [41:44] It's actually funny, the most important [41:46] This is actually so important. I I can't [41:48] believe I missed this. The most [41:49] important thing in any video you make [41:52] doesn't matter if it's a shitty Loom or [41:54] a shitty like little video or your [41:55] iPhone is audio quality. If your audio [41:58] is bad, I'm bouncing. I've watched some [42:00] some demos that are like 720p and just [42:04] like the guy's in like a weird room in a [42:06] basement, but his audio is crisp. And [42:08] what he's working on is actually [42:10] genuinely interesting, so I stick [42:11] around. And so I don't care about the [42:13] the rest of it. I don't care what he [42:13] looks like, anything. I'm like, I can [42:15] hear him and I can see what he's doing. [42:17] That's it. Yeah. And so I think if every [42:19] founder could just invest in one of the [42:21] super cheap DJI mics, find them used [42:24] online, [42:25] or just any mic that's not your MacBook [42:27] mic, [42:28] your or your AirPod mics, those don't [42:30] count, [42:31] that would be really good. And then for [42:33] editing on my actual laptop I use [42:35] DaVinci. Yeah. [42:37] And maybe the most important part of my [42:38] stack he's not here right now, but [42:40] Vikrant. Vikrant has made all the Hey [42:42] Clicky videos. And without that guy [42:44] making me look good and making us look [42:46] good, none of this would happen. And so [42:48] I don't want to sit here and be like I [42:49] made all all videos and did all this [42:50] stuff. Every video we just looked at, [42:52] apart from the the Bad Zip School ones [42:53] which were made by me, are made by [42:55] Vikrant. And so, having a person that is [42:57] actually talented at filmmaking, [42:59] storytelling, writing next to you is [43:01] like the biggest [43:03] bonus a founder could have today. So, a [43:05] big part of how I work with Vikrant, [43:07] who's the person who helps me with all [43:08] these videos, is [43:10] we just brainstorm. We're like, "Hey, we [43:12] made this feature. What should How [43:14] should we make the video now? [43:16] What are we thinking? Where should we do [43:17] it?" And so, Vikrant is actually the [43:19] main guy that's like, "Hmm, I had this [43:21] shot idea." Or, "Hmm, I had this hook [43:23] idea." Or, "Hey, I have this other [43:25] idea." And often I'm the guy that's [43:27] bringing the script and the song and the [43:29] vibe that I'm going for. [43:31] Cuz as the founder, I think you'll [43:32] slowly find what are the things that [43:34] you're good at in video making. There's [43:36] some aspect to it. So, for me, I find [43:38] I'm really good at writing and I'm [43:39] really good at the music. [43:41] And then Vikrant is good at literally [43:43] everything else. From shooting to [43:45] picking the set to editing after to, you [43:48] know, all the little motion graphics he [43:50] does. Just so talented. And then as the [43:52] founder, me as Farza, I can go back to [43:54] focusing on growth, hiring, making the [43:57] product. I don't want to spend [43:59] all my time on making videos, you know? [44:01] I don't think any founder watching this [44:02] wants to do that. And and look, I think [44:04] a lot of these a lot of people watching [44:05] probably can't afford to hire a [44:07] filmmaker or like an actual person on [44:08] their team. They're maybe just solo or [44:10] they're co-founder. [44:11] And what I'll say is, "Hey, go back to [44:13] my old Zip School videos. Go back to [44:15] Just scroll through any of my social [44:17] media and look at all my old shitty [44:18] videos and just start somewhere. That's [44:20] the most important thing about making [44:23] launch videos and launching today is you [44:25] got to start. And the more you keep [44:27] doing it every day, [44:28] the more natural it'll become and slowly [44:31] hopefully more people check out your [44:32] product. Hopefully you can then hire [44:34] someone that does it even better than [44:35] you. So, for example, I'm not the [44:37] engineer at my company anymore. I'm not [44:38] the core engineer. I have someone that's [44:40] a lot better than me. [44:41] >> Yeah. [44:41] >> And same with filmmaking, same with [44:42] design, same with every aspect of my [44:43] company. But in the early days as a [44:45] founder, [44:47] you're doing everything. [44:47] >> it all. [44:48] >> You're doing taxes, you're doing design, [44:49] you're doing filmmaking. And so, you got [44:51] to do everything. You can't be a little [44:52] baby about it. You got to take out your [44:54] camera and start posting some cringe [44:55] stuff. [44:56] Cuz in order to not be cringe, you must [44:58] first be cringe, so [44:59] Yeah. And then yeah, for all of my [45:01] writing, I always use this thing called [45:03] Free Write. Um it's free to download, [45:04] freewrite.io. It's my app. I built it [45:07] actually just for this purpose. [45:08] >> Mhm. [45:08] >> Uh yeah, free, open source, you can [45:09] check it out. Maybe this is one thing I [45:11] should I didn't mention at all before is [45:12] writing is so important. So spend as [45:14] much time possible before the video uh [45:17] writing it. So before you get in front [45:18] of the camera and maybe record it, if [45:20] you've never made a video, [45:22] don't worry about this, but if you're [45:23] like on your 15th, 20th, 30th video, [45:26] try writing. Try writing for just 10 [45:27] minutes about what the what it should be [45:29] and then record it. Um you'll find that [45:32] the more you think through it before you [45:33] get to the video, uh the better it can [45:36] possibly get. Cuz then you can catch a [45:37] bad idea before it gets here. [45:39] >> Yeah. [45:39] >> And that's it. That's that's that's [45:40] that's my stack. [45:41] >> I think you you have a challenge for [45:42] everybody who's watching for something [45:44] they can do today to get started, right? [45:46] >> My challenge to everyone watching this [45:48] is um [45:50] if you've never really made videos or if [45:52] you uh have never really talked about [45:54] your work online, but you've been [45:55] hacking, you've been tinkering, you've [45:56] been working on something, [45:58] super simple. Every day, just spend 10 [46:01] minutes, take your phone out, iPhone or [46:03] Android, it's okay, and then just record [46:05] yourself talking about what you were [46:06] doing to what you did that day, and then [46:08] post it to your social media of choice, [46:10] Instagram, TikTok. You know what? Make [46:12] it Make a new Instagram account. So your [46:14] your existing followers don't even know. [46:15] You do just post it somewhere. And do [46:18] that for 30 days straight, and I promise [46:21] you, within those 30 days, one of those [46:23] videos will surpass probably 1,000 [46:25] views. And then you can look back and [46:27] ask yourself, [46:29] hm, that one did 1,000 views. [46:31] Maybe the next one, and you can learn [46:32] from that, and you can make the next one [46:35] uh based off the one that got 1,000 [46:36] views. And you can slowly get better and [46:38] better and better. That's the challenge, [46:39] yeah. 30 days straight, every day one [46:41] video. Send it to me. Uh I'll be I'll be [46:44] I'll be one of your likes [46:45] uh if you're doing this challenge, and [46:47] yeah, go for it. [46:47] >> And it's never been easier to do this, [46:49] right? To get that message out. [46:50] >> It's never been easier. It's never been [46:51] easier. [46:52] >> Farza, thank you so much for being here. [46:54] This is incredible advice. [46:55] >> Anytime, man. [music] [47:00] >> [music]