How I Hit 10M Users in Just 1 Year l Creati, Ella Zhang

EO13:24Added Aug 31, 2026

Ella Zhang is the Founder & CEO of Creati, a fast-growing AI video generation tool built for marketers. In just one year, Creati surpassed over 10 million downloads and reached $10M in revenue.

Watch on YouTube →
Contributed by 刘嘉琪

Transcript

Transcript format
  1. Intro

  2. 00:00To be honest, like our first product
  3. 00:02looks really rough. We actually only
  4. 00:04used two weeks. One day we have over
  5. 00:07like 100 people just flowing this
  6. 00:09product. Our users actually just you
  7. 00:12know like doubled. There is huge
  8. 00:14increase in the user and there's huge
  9. 00:16increase in the revenue. That is a time
  10. 00:18we know okay we hit something right. We
  11. 00:20have over 10 million downloads in a year
  12. 00:23and we now have over 10 million
  13. 00:25annualized revenue within a year. It's
  14. 00:27not just about fantastic UI, UX or how
  15. 00:30you're marketing it, right? It's more
  16. 00:32like what result users are get. You
  17. 00:35don't need to wait for the product to be
  18. 00:37perfect to launch it. Most of the time
  19. 00:39in the AI era, things will change every
  20. 00:41day. If you wait too long, you may never
  21. 00:44be able to launch your product. So, you
  22. 00:46just need to launch it, develop it,
  23. 00:49learn from it at the same
  24. 00:54time. Hi, I'm Ella. I'm the founder and
  25. 00:57the CEO of CRY AI. Criti is the agentic
  26. 01:01AI to augment your marketing video at
  27. 01:03scale. Just drop one product image and
  28. 01:06Criti write your script, shoot your
  29. 01:08video and even do the processing for
  30. 01:11your marketing video without the need of
  31. 01:14full media crew. Cringy is growing
  32. 01:16pretty fast. We have over 10 million
  33. 01:19downloads in a year and we now have over
  34. 01:2110 million annualized revenue within a
  35. 01:24year thanks to our network effect
  36. 01:26between the influencer and the
  37. 01:31[Music]
  38. Tip1: Apple's Rule: Talk to Your Users in Person

  39. 01:33brands. I started my career as actually
  40. 01:36a hardware engineer at Apple. We are
  41. 01:38building the first generation of
  42. 01:40AirPods. We don't even know what product
  43. 01:42people actually need. So we invite a lot
  44. 01:46of users to our small interview room and
  45. 01:48we ask them a lot of questions like what
  46. 01:51do you need actually for the earpods?
  47. 01:53What do you actually have the pinpoints
  48. 01:56for? Uh things like that. At Apple when
  49. 01:58we conduct the interviews instead of
  50. 02:00just doing online surveys we actually
  51. 02:03more prefer to invite the users to the
  52. 02:05interview room and conduct like
  53. 02:07face-toface interviews. Right? When you
  54. 02:09ask people to fill out the survey
  55. 02:10online, people most of time they just,
  56. 02:12you know, click the first one, whatever
  57. 02:15they see there, right? Because they want
  58. 02:17to just jump to the end. No one really
  59. 02:20want to get their feedback. So you may
  60. 02:22get the wrong answers. And another more
  61. 02:24important thing actually is when you are
  62. 02:26talk to your users face to face, you
  63. 02:29actually get more than what you want to
  64. 02:31ask, right? Like people they always have
  65. 02:34some assumption that okay maybe user
  66. 02:36need A b and C but actually user don't
  67. 02:39care about them like at all right like
  68. 02:42they only care about what they care
  69. 02:44about like D but you don't have that on
  70. 02:46your survey so you'll get the wrong
  71. 02:48answer that's the reason actually Apple
  72. 02:50prefer to have face-to-face user
  73. 02:52interview when they are building a new
  74. 02:54product and one important thing we learn
  75. 02:56at Apple is you know you really need to
  76. 02:59understand people's pain points you need
  77. 03:01to really know why people need it when
  78. 03:04you are building it. We have our like a
  79. 03:06lot of fantastic ideas or imagination
  80. 03:10but those are not solving the pain
  81. 03:12points of real customers and that's not
  82. 03:14we want right so meeting people or
  83. 03:17knowing people's pain points actually is
  84. 03:19the first step when you need to create a
  85. 03:21great product. The AirPods project is
  86. 03:23like it's amazing uh experience for me.
  87. 03:26Many people said I should stay there for
  88. 03:28longer. But the thing is that I always
  89. 03:30pursue for building something new or
  90. 03:33building something I think I feel most
  91. 03:35excited for. When I was a child, I found
  92. 03:38I get information much easier when I get
  93. 03:41it in the form of video instead of like
  94. 03:44long text or like blog post or things
  95. 03:46like that. So it is always my dream that
  96. 03:49in the childhood that if one day we
  97. 03:52could turn a lot of communication the
  98. 03:55learning the information gathering just
  99. 03:57in the video format that will be really
  100. 04:00helpful for a lot of people sometimes
  101. 04:02you know it's it's pretty pretty hard
  102. 04:04for people from the giant company to
  103. 04:07actually quit and really doing the
  104. 04:08startup when I looking back I really
  105. 04:11miss the time there at big company
  106. 04:14enjoying with my friends and my family
  107. 04:16Still I think I don't regret uh because
  108. 04:19I am doing what I really passionate
  109. 04:22about and I believe that will help a lot
  110. 04:25of people that is more important to
  111. Tip2: Done Is Better Than Perfect

  112. 04:31me when I saw the generative AI even
  113. 04:34it's back in 2020 I believe that one is
  114. 04:38will change the world and it will be the
  115. 04:40future because at that time I saw a lot
  116. 04:43of people they spend over 80% of their
  117. 04:46time just online instead of offline,
  118. 04:48right? So there is a huge trend that
  119. 04:50people actually are shifting from
  120. 04:51offline to online and then people will
  121. 04:54definitely need much much higher demand
  122. 04:56for the great content online. But on the
  123. 05:00offset that I saw it's pretty pretty
  124. 05:02hard for people to create great content
  125. 05:04right the requirement for great content
  126. 05:07actually growing higher and higher. The
  127. 05:09supply of the great content actually uh
  128. 05:12have a huge gap between the supply and
  129. 05:14the demand. I'm pretty certain that you
  130. 05:16know it's only the technology which
  131. 05:18could like merge the gap and help people
  132. 05:21with that. That's a time that I think
  133. 05:23okay even though I am not experts I
  134. 05:26don't know a lot about software or I
  135. 05:28don't know a lot about AI but you know I
  136. 05:31can always learn once I find the
  137. 05:33motivation I would know what I want to
  138. 05:35do and I can shift to that and I can
  139. 05:38learn that pretty quickly to be honest
  140. 05:40like our first product looks really
  141. 05:43rough. We don't have really fantastic UX
  142. 05:47or UI. We even don't have the batch mode
  143. 05:50in the beginning. But that's totally
  144. 05:52okay. You know, like sometimes when the
  145. 05:54product is really rough, but users
  146. 05:56still, you know, like willing to use it,
  147. 05:59that means something. If you have a
  148. 06:01really like bad design product, but
  149. 06:04people are still using it. They're like,
  150. 06:06okay, there's some real value thing in
  151. 06:08that one. It's not just about fantastic
  152. 06:11UI, UX, or how you're marketing it,
  153. 06:13right? It's more like what result users
  154. 06:16are get whether they can use it to get
  155. 06:18the benefit they want. You don't need to
  156. 06:21wait for the product to be perfect to
  157. 06:23launch it because you know like most of
  158. 06:25the time in the AI era like things will
  159. 06:27change every day. If you wait too long
  160. 06:30you may never be able to launch your
  161. 06:32product. So you just need to launch it,
  162. 06:34develop it, learn from it. At the same
  163. Tip3: Don't Chase Trends, Chase User Needs

  164. 06:41time, there's one lesson I learned as a
  165. 06:44founder or I will say the most painful
  166. 06:46lesson I learned. Our first product
  167. 06:48actually is called ZMO.ai. At the time I
  168. 06:51was building Zmo. I'm not sure like why
  169. 06:54I'm building this like why people need
  170. 06:56this. The thing is at the time me
  171. 06:58journey is super hot. Everyone is
  172. 07:00chasing that right. So just like so many
  173. 07:03other founders, I'm like, "Okay, text to
  174. 07:06image is so popular. We definitely need
  175. 07:08to enter this domain. So let's do it."
  176. 07:10But I don't really know why people are
  177. 07:12generating those images. So I don't
  178. 07:15really understand the users need. When I
  179. 07:17was building the product, we actually
  180. 07:19don't know what the business model will
  181. 07:21be and we even don't know what the next
  182. 07:23step will be, right? So the revenue
  183. 07:26actually doesn't go that high as the
  184. 07:28users grow, right? and we suddenly just
  185. 07:31get lost right we don't know like why
  186. 07:33people are using it like why some people
  187. 07:36pay why some people don't pay for it
  188. 07:38when we have this difficulty the only
  189. 07:40thing I can think about is talking to
  190. 07:42your customers right that is a thing I
  191. 07:44learned from Apple we do email our users
  192. 07:48or most active users email them like
  193. 07:51over 500 and finally we get 300 actually
  194. 07:54we have the face toface conversation
  195. 07:57with them because I believe talk to the
  196. 07:59users is the only way we can find
  197. 08:02whether there is a true need or what do
  198. 08:05they really need after talking to more
  199. 08:07than like 300 users we find that a lot
  200. 08:10of people they create the images it's
  201. 08:13just out of curiosity the AI is so new
  202. 08:15to them they just want to try it out but
  203. 08:18they do a small group of people they
  204. 08:21they actually do have the need for it
  205. 08:23not designers they don't know how to
  206. 08:25create all those cool images they do
  207. 08:28need actually those content for
  208. 08:30themselves for their company uh for
  209. 08:33marketing. So we're like okay we finally
  210. 08:36find people who have real needs for
  211. 08:38that. Most of them who are on Amazon or
  212. 08:41eBay or Etsy they actually use their
  213. 08:43phone directly to take the photos of
  214. 08:45their product. So they hope that okay
  215. 08:48the social media is also on the phone
  216. 08:51photos and the videos are also on the
  217. 08:53phone. it will make him much more sense
  218. 08:55if you just build the product on the
  219. 08:57model. After talking to them, we finally
  220. 08:59made the decision. Okay, we will
  221. 09:02reposition this product and we will
  222. 09:04build it for people who really have the
  223. 09:06needs which is marketing and we will
  224. 09:08also shift it actually from the PC side
  225. 09:10to mobile. That's the time when the
  226. 09:12pivot is happening.
  227. Tip4: The Growth Flywheel Behind $10M ARR

  228. 09:18We are building the mobile version of
  229. 09:20creati in a way that it's not actually
  230. 09:23just a tool but it's more like a
  231. 09:26community for both influencers and the
  232. 09:28brands. The reason actually we have
  233. 09:31features or templates for both
  234. 09:32influencers and brands is actually we
  235. 09:35find if some brands they want to do
  236. 09:38marketing it's impossible actually to
  237. 09:41avoid uh the influencers contribution
  238. 09:43because influencers they are the people
  239. 09:46who actually know best about the social
  240. 09:49media who actually know much better
  241. 09:51about what content is more engaging and
  242. 09:54more attractive to users. So when we are
  243. 09:57building creat from day one we know okay
  244. 10:00we need to create the actually the
  245. 10:03network effect between the influencers
  246. 10:06and the brands they they know the
  247. 10:08internet so well after they create the
  248. 10:10template they will also will post those
  249. 10:12videos uh on their own channel on the
  250. 10:15social media and if that video templates
  251. 10:17actually is a great one that video will
  252. 10:20actually go viral it will actually get
  253. 10:22millions or tens of millions of views
  254. 10:25right and Those viral videos actually is
  255. 10:28the best marketing strategy for us to
  256. 10:30acquire more and more brands because
  257. 10:32when the video go viral that will bring
  258. 10:34a lot of lot of users to the
  259. 10:38createl the end of funnel there will be
  260. 10:40the brands that will really use this
  261. 10:42video templates to create their engaging
  262. 10:44ads video. So the brands actually when
  263. 10:46they are using those brands as video
  264. 10:48they will pay more uh to create this
  265. 10:50platform so that we can hire more
  266. 10:52influencers in the community and the
  267. 10:54influencer can create more viral video
  268. 10:56templates for the brands. So it's
  269. 10:59actually a very nice flywheel actually
  270. 11:01we're building between the influencers
  271. 11:04and the brands and we believe that is
  272. 11:06the key differentiation CR is have
  273. 11:09compared to all other like AI marketing
  274. 11:11tools is because we believe the
  275. 11:13ecosystem actually need both sides. We
  276. 11:16need both influencers and brands to help
  277. 11:19them to grow marketing at scale.
  278. Tip5: Adapt Fast. Users Are the Signal

  279. 11:26The AI world just changed daily. Every
  280. 11:28day you wake up, there will be new
  281. 11:30technology and new product, right? And I
  282. 11:33think building in the AI era is quite
  283. 11:36different from like 20 to 30 years ago
  284. 11:39when people are just enter the mobile
  285. 11:41boom. The thing is actually I believe
  286. 11:45maybe 99% or maybe all the products that
  287. 11:48you saw today maybe not the final
  288. 11:51version of what they will be at the end
  289. 11:54of AI evolution. The requirement for the
  290. 11:56founder in the AI era for AI startup
  291. 11:59actually is kind of different from
  292. 12:02previously because those founders really
  293. 12:04need to adapt really quickly and
  294. 12:07flexibility to those changes. And one
  295. 12:10thing we think actually is helping us to
  296. 12:13adapt to all those changes is we talk to
  297. 12:16our users every day and to learn what is
  298. 12:20the key thing that won't change even if
  299. 12:23the technology change. So the thing that
  300. 12:25we find out you know is what kind of
  301. 12:28content will go pretty like viral on the
  302. 12:31social media is all the things that our
  303. 12:34users or our customers care about. Even
  304. 12:36though technologies change, the
  305. 12:38important thing is once you get those
  306. 12:41industry insider data and knowledge uh
  307. 12:44for that vertical that will build your
  308. 12:46mode and that will actually be your key
  309. 12:49value along with those all technology
  310. 12:51changes daily or those technology change
  311. 12:54will just help you instead of you know
  312. 12:56like disrupted to your product.