How Did Creati Founder Ella Zhang Reach 10M Users in Just 1 Year?

TL;DR

Ella Zhang says Creati reached over 10 million downloads and over 10 million in annualized revenue within a year by delivering useful marketing-video results, launching an imperfect product quickly, learning directly from users, and benefiting from a network effect between influencers and brands. The first product took only two weeks to build, showing why its launch, interview, and product-value lessons are worth examining closely.

Transcript

To be honest, like our first product looks really rough. We actually only used two weeks. One day we have over like 100 people just flowing this product. Our users actually just you know like doubled. There is huge increase in the user and there's huge increase in the revenue. That is a time we know okay we hit something right. We have over 10 mill... Read More

Key Insights

• Pain points come before features: Zhang’s Apple experience taught her to begin by understanding why people need a product. Teams may have impressive ideas and detailed assumptions, but those ideas can fail to address real customer problems. Meeting users and identifying their actual pain points is therefore presented as the first step toward creating a product that people will choose to use.
• Interviews expose missing assumptions: A prepared survey only asks about options its creator has already imagined. Zhang explains that a team might ask whether users need A, B, or C even though users care only about D. Face-to-face conversation can surface that unlisted priority, giving the team information that a fixed survey could never capture and preventing development around the wrong assumptions.
• Survey convenience can distort answers: According to Zhang, people completing online surveys may click the first option they see because they want to reach the end. That behavior can produce answers that do not represent what users genuinely want. Apple’s preference for in-person interviews was therefore not merely about collecting more feedback, but about improving the depth and reliability of what the team learned.
• Personal experience shaped the mission: Zhang’s preference for receiving information through video began when she was a child. She found video easier to understand than long text or blog posts and imagined communication, learning, and information gathering becoming more video-based. This long-standing interest, rather than a newly discovered market trend alone, helped define the problem she wanted her company to address.
• Passion justified leaving Apple: The first-generation AirPods project was an experience Zhang describes as amazing, and she missed the time spent at a large company with friends and family. Even so, her desire to build something new and exciting outweighed the appeal of staying. She frames the decision around doing work she cares about and believes will help many people.
• Online behavior created demand: Zhang’s opportunity thesis began with the observation that many people spent over 80% of their time online. She expected this shift from offline to online activity to produce much greater demand for high-quality digital content. The opportunity was not simply that video was popular, but that online audiences increasingly required more strong content than creators could readily supply.
• Generative AI could close the gap: Zhang saw a mismatch between rising requirements for great online content and the difficulty of producing it. She became convinced that technology could merge the gap between content supply and demand. Her belief in generative AI’s future dates to 2020, when she decided its potential mattered more than her lack of existing expertise in software or AI.
• Motivation enabled rapid learning: Zhang openly says she was not an expert and did not know much about software or AI when she identified the opportunity. Her response was to treat those limitations as learnable rather than permanent. Once she found a strong motivation and knew what she wanted to build, she believed she could shift fields and acquire the necessary knowledge quickly.
• Rough usage signals real value: The first product lacked strong UI, polished UX, and batch mode, yet users were still willing to use it. Zhang interprets that behavior as unusually meaningful because poor design creates friction. Continued adoption under those conditions suggests that the outcome delivered by the product is valuable enough for users to tolerate an unfinished experience.
• Outcomes matter beyond presentation: Zhang rejects the idea that success depends only on attractive UI, refined UX, or marketing. Her central test is whether users obtain the result and benefit they came for. The product’s appearance can remain rough during its earliest stage if its core function already solves a meaningful problem and produces an outcome users value.
• Early traction revealed product fit: More than 100 people entered the product in one day, after which the user count doubled and revenue rose substantially. Zhang describes this combined movement as the point when the team knew it had hit something right. The signal came from observable user and revenue changes, not from internal confidence about the idea or the interface.
• AI speed rewards simultaneous action: Zhang says conditions in the AI era change every day, making a long pursuit of perfection dangerous. A team that waits too long may never launch because the environment keeps moving. Her alternative is a concurrent cycle: release the product, continue developing it, and learn from actual usage at the same time rather than treating those as separate phases.

Questions & Answers

Q: How did Creati founder Ella Zhang reach 10 million users in one year?

Zhang says the company launched a rough product quickly and focused on the results users received rather than waiting for perfect UI or UX. The first version took only two weeks, yet people continued using it despite missing features such as batch mode. A day with more than 100 people flowing into the product was followed by doubled users and major increases in usage and revenue, signaling that the product delivered real value. Growth was also supported by a network effect between influencers and brands. Within a year, the company recorded over 10 million downloads and over 10 million in annualized revenue.

Q: What does Creati do for marketing videos?

Zhang describes the product as agentic AI that augments marketing-video production at scale. A user can provide one product image, after which the product writes a script, shoots the video, and handles video processing. The purpose is to produce a marketing video without requiring a full media crew. This addresses the difficulty of creating enough strong online content as demand continues to grow.

Q: Why did Zhang prioritize face-to-face user interviews?

Face-to-face interviews allow a team to learn more than its prepared questions anticipate. Zhang says online survey respondents may click the first answer they see simply to finish, which can generate misleading results. A survey might offer A, B, and C while the user actually cares about an unlisted D. Direct conversation reveals those unexpected priorities and helps the team understand both the pain point and the reason the customer needs a solution.

Q: What did Zhang learn while working on AirPods at Apple?

Zhang worked as a hardware engineer on the first generation of AirPods. Because the team did not initially know what product people needed, it brought users into a small interview room and questioned them about their needs and pain points. She learned that attractive ideas and internal imagination are insufficient when they do not solve real customer problems. Her central lesson was that meeting people and understanding why they need something should come before building the product around assumptions.

Q: Why did Zhang leave Apple to start something new?

Zhang says she has always pursued opportunities to build something new or something she finds especially exciting. Her childhood preference for learning through video gave her a lasting interest in moving communication, learning, and information gathering into video formats. Leaving a large company was difficult, and she still missed time there with friends and family. Nevertheless, she does not regret the choice because she is pursuing work she cares about and believes can help many people.

Q: Why did Zhang pursue generative AI in 2020?

In 2020, Zhang believed generative AI would change the world and become the future. She observed that many people spent over 80% of their time online, creating greater demand for excellent online content. Producing that content was still difficult, leaving a large gap between supply and demand. She believed technology could close that gap, and that conviction motivated her to learn software and AI despite not being an expert in either field.

Q: Why launch an AI product before it is perfect?

Zhang says AI changes every day, so waiting for perfection can delay a product until it never launches. Releasing early allows development, learning, and real usage to happen at the same time. Her own first product took two weeks and lacked polished UI, UX, and batch mode. Because users still chose to use it, the team gained evidence that the underlying result mattered, even before the overall experience was refined.

Q: What showed Zhang that the product had real value?

The earliest signal was that users tolerated a rough product with weak design and missing functionality. Zhang reasons that people will endure those shortcomings only when the product gives them a benefit they genuinely want. The stronger signal arrived when more than 100 people flowed into the product in one day, users doubled, and revenue increased substantially. Together, continued usage and rapid growth showed that the product’s output mattered more than its unfinished presentation.

Summary & Key Takeaways

• Starting with direct user research: Ella Zhang began her career as a hardware engineer at Apple, working on the first generation of AirPods. Because the team did not yet know what people needed, it invited users into a small interview room and asked about their needs and pain points. Apple preferred face-to-face interviews over online surveys because respondents might simply click the first available survey answer, while a conversation could reveal important needs that the team had never thought to include.

• Choosing a personally meaningful problem: Although people encouraged Zhang to remain at Apple longer, she wanted to build something new that genuinely excited her. Since childhood, she had found information easier to absorb through video than through long text or blog posts. That experience shaped her ambition to move more communication, learning, and information gathering into video. Leaving a large company was difficult, but she says she does not regret pursuing work she feels passionate about and believes can help many people.

• Recognizing the online content gap: Zhang saw generative AI in 2020 as technology that could change the world. She observed that many people spent over 80% of their time online rather than offline, increasing demand for strong online content. At the same time, creating great content remained difficult, producing a substantial gap between supply and demand. Although she was not a software or AI expert, identifying that gap gave her the motivation to change direction, learn quickly, and pursue a technology-based solution.

• Launching a rough first product: The initial product was built in only two weeks and lacked polished UI, UX, and even a batch mode. Zhang considered continued usage despite those limitations an important signal. If people willingly use something with poor design, she argues, they are probably receiving real value from its results. Rather than judging the product primarily through appearance or marketing, her team focused on whether users obtained the benefit they wanted and treated the early launch as an opportunity to develop and learn simultaneously.

• Finding rapid growth and momentum: At one point, more than 100 people flowed into the product in a single day, users doubled, and both usage and revenue rose substantially. That was the moment Zhang felt the team had found something right. She says the company achieved over 10 million downloads and over 10 million in annualized revenue within a year. She attributes the fast growth partly to a network effect between influencers and brands, while emphasizing that rapid change in AI makes waiting for perfection especially risky.