What Comes After a Viral Demo: Farza’s Case for Launching as a Continuous Practice

1. Guest Background

The supplied guest background identifies the interviewee as Barza Majid, the founder of Hey Clicky and a previous founder of BuildSpace. During the conversation, he is generally called Farza. BuildSpace helped people work on their own ideas; Hey Clicky explores a new interface for AI. Its central character, Clicky, is a small blue companion that follows the computer cursor and is described as an AI teacher capable of guiding users as they work.

That product makes demonstration particularly relevant. A description of a cursor companion leaves important questions unanswered: what does it notice, how does it guide someone, and what happens on the screen? The interview examines how Farza presents such interactions to prospective users, then decides whether the attention represents meaningful interest. It is an analysis of a founder’s working practice, rather than an independently audited account of acquisition performance.

Published by Y Combinator, the episode runs for approximately 47 minutes. The host refers to Farza’s more than a decade of experience publishing online videos and discusses video as an important way he reached audiences at BuildSpace and Hey Clicky. That experience matters when assessing the advice. His apparent ease on camera sits behind years of practice, and the conversation also acknowledges production collaborators. Readers should distinguish the accessible starting point he recommends from the accumulated judgment and assistance visible in the successful examples.

2. What the Episode Covers

Farza begins by changing the unit of a launch. In his account, the environment around 2020 encouraged founders to organize around community sites, Product Hunt, Reddit, and press coverage. He describes algorithmic platforms as allowing a continuing stream of smaller demonstrations. His team publishes demos almost daily, with a more polished launch occasionally, perhaps once a quarter. This is a description of his operating rhythm; the interview does not establish an optimal schedule for other businesses.

The rhythm reaches back into development. While discussing a feature, the team also considers the Instagram caption, the opening of a video, and the interaction that will eventually be shown. The host calls this demo-driven development. Communication becomes part of defining the work rather than a packaging exercise after implementation. Farza connects that practice to a commercial objective: a channel that can repeatedly bring people into the product. Internet recognition is insufficient if nobody uses what the team has built.

Clicky’s first demonstration provides the episode’s most useful connection between distribution and product learning. Farza says it began as a side project that he open-sourced because he doubted anyone would care. He recorded the initial video roughly five minutes before dinner. About six months later, the project had become his company. The host describes the video as approaching three million views, but Farza directs the discussion toward approximately 7,000 downloads and around 100 people who used the product every day for five consecutive days.

He says he contacted every one of those users and discovered uses he had not anticipated. Their behavior and conversations helped explain why he continued. The distinction is consequential: exposure opened a door, while repeated use supplied a reason to investigate further. The episode offers an early signal of interest, without establishing longer-term retention.

Other examples show how a fuller narrative can introduce the product. One video moves from the founder’s recent experience into Clicky guiding color grading in DaVinci Resolve. Farza reports 21 million Instagram views, continuing traffic, and subsequent use as a Facebook advertisement. He says it took four to six hours to make. Those details describe a reusable piece of content, although the episode does not disclose its advertising economics.

The conversation also preserves the less glamorous practice behind these examples. Farza revisits a ZipSchool video with 23 likes and a rudimentary landing-page announcement made when he had about 80 followers. He treats both as legitimate launches and opportunities to learn product communication. Team updates offer another format: explaining agent orchestration, control and pause improvements, and a fix for Clicky’s following behavior. He says these updates take about ten minutes to make, now appear every two weeks, and began as public accountability. One reportedly approached a million Instagram views. Together, these formats make launching a recurring part of work with several levels of production effort.

3. Core Views: Reasoning, Examples, and Limits

The strongest creative principle in the interview is that attention begins with recognition. Farza places prospective copy inside a tweet interface and asks whether he would stop to read it. In the first Clicky demo, Xcode was a deliberate opening frame intended to interest curious developers. Magnification and movement made the tiny assistant easier to notice. These are concrete decisions about audience and visibility: a viewer needs to recognize the setting and see the relevant action before a product explanation can work. The founder’s own response is still a preliminary test, rather than proof that customers will respond similarly.

Preparation serves the same purpose. For a roughly one-minute video, Farza recommends explaining the work and its importance within about ten seconds, with setup occupying approximately 10–20% of the runtime. His screen-aware dictation example makes that advice substantive. The demo first acknowledges an ordinary capability that competitors already offer: speech appearing as text at the cursor. It then introduces his personal difficulty with knowing what he wants to write. That difficulty creates a reason to show the additional capability. Crucially, he confines the problem to his experience instead of claiming that everybody shares it.

The apparent spontaneity was prepared. Farza says he wrote that video line by line three days beforehand and mentally rehearsed it around ten times. He nevertheless thought it was too poor to edit; a collaborator pushed it through, and he reports roughly three million Twitter views. This supports two separate observations: rehearsal can give a speaker something clear to say, and embarrassment can be a poor publication filter. It does not isolate rehearsal as the cause of the result. The episode contains no controlled comparison separating script, product novelty, audience, and distribution.

A second example shows how actions themselves can carry a story. Using GPT real-time capabilities to generate shell commands, Clicky plays music in Spotify, lowers the volume to 50%, and opens a Stripe dashboard. Farza says the video has no cuts and reached 3.8 million views on Twitter and 4.6 million on Instagram. The host identifies beats despite the continuous recording: the hands-free premise is introduced, demonstrated, and extended. Each completed action creates a reason to anticipate the next. The dashboard’s mention of $140 in a day belongs to the demonstration; it establishes no broader revenue performance. Nor does an uninterrupted example establish reliability across other tasks or conditions.

Distribution requires a different standard of judgment from storytelling. Farza describes Twitter as historically a particularly poor conversion channel for him, including about 2.9 million views on the dictation video followed by very low account creation. He contrasts that with a YouTube video receiving around 20,000 views and, he says, approximately 20% conversion from viewing to downloading, on a channel with 872 subscribers. He also estimates that Instagram supplied 70% of traffic. These statements mix registrations, downloads, and traffic share. Without complete attribution methods and measurement windows, they support investigating channel differences, not adopting his percentages as benchmarks or treating traffic as revenue.

The design-agency example gives targeting a more modest scale. Farza says about three agencies discovered the product naturally. The team then made TikToks explicitly addressing agency founders, averaging roughly 4,000 views and generating about 50 leads. His explanation is that the audience was more relevant. That is plausible within his account, but the interview reports no resulting deals. The example supports testing whether business buyers can be reached on consumer platforms; it does not settle the relative economics of that approach.

Finally, Farza treats production as a strategic capability. He names a filmmaker and an engineer among his earliest hires and says Clicky once had three people making videos and one engineer. His rationale is that audiences and communities are harder to copy than software features. That is his allocation judgment, with substantial opportunity costs for another team. His claim that a direct founder demo may outperform a $20,000 production also lacks comparison data. He acknowledges that the episode shows successful work and that much earlier work failed. A drawing demo optimized for Twitter reportedly performed better unchanged on Instagram. Taken together, these qualifications favor experimentation over a fixed formula for platform, format, or staffing.

4. Learning and Application

A useful starting point is a limited commitment to showing actual work. Farza recommends a roughly 30-second phone video and proposes spending about ten minutes each day for 30 days. Treat that as a practice schedule: choose a real improvement, explain its use, demonstrate it, publish, and return to building. His prediction that a video may exceed 1,000 views is not a validated guarantee. The exercise should produce material to learn from even when that prediction fails, provided the founder reviews what was understandable and who responded.

Bring a small communication test into feature discussions. Draft a caption, an opening, and the final interaction before committing to a complicated presentation. This applies demo-driven development as a clarification tool: if the team cannot identify the user and the useful action, the proposed benefit may need further thought. Keep subsequent use in the decision. A feature that is easy to demonstrate can attract attention without becoming valuable in repeated work. Clicky’s continuing users and the conversations with them show where to seek stronger evidence about what deserves further development.

Increase preparation as experience accumulates. Farza suggests trying ten minutes of writing before recording once someone has made around 15–30 videos. FreeWrite, the free and open-source application he built for writing, is one possible tool rather than a prerequisite. A short draft can expose a missing connection between problem and demonstration before filming becomes expensive. Check the opening frame as well: can the intended viewer recognize the setting and see the important action? Studying where one personally leaves other videos can suggest improvements, but those observations still need testing with an actual audience.

Allocate production resources to comprehension first. Farza’s toolkit spans screen recording, phones, cameras, microphones, and editing software, yet he gives clear audio priority over image quality. His willingness to watch an interesting, intelligible 720p demo supports checking sound and screen readability before buying equipment. He also credits his filmmaking partner for successful production, describing a division in which he contributes script, music, and overall feeling while the partner handles shooting, locations, editing, and motion graphics. A team can use that distinction to identify a bottleneck. Hiring becomes a capacity decision when recurring work justifies it; founders without the budget can begin with basic self-produced demonstrations.

Measurement should follow the steps people actually take. Use distinguishable channel entry paths, as Farza suggests, and record viewing, registration, download, and subsequent use separately, with explicit observation periods. For phone visitors, emailing a link to themselves can be a useful intermediate signal, but it is not installation or retention. Where a customer group appears organically, make a focused demonstration for that role and track the quality of the resulting leads through to sales. The agency example provides a reason to try this sequence, without supplying a forecast of closed business.

Use public updates when there is meaningful progress to explain. A fix, a control improvement, or an orchestration change can support a direct account of what changed and why, without requiring a cinematic launch. Farza’s emphasis on product work rather than founder lifestyle provides an editorial boundary. The aim is to help people understand and assess the product. Review who arrived, who tried it, who continued, and what their feedback changes about the next development decision.

Source

• Original episode: This Is How You Should Actually Make Launch Videos