Steven Cravotta
Founder of Juiced Media (organic TikTok agency), creator of Puff Count and the original Wordle iOS app
Where to hear and read about Steven Cravotta1 episode or test
Steven Cravotta runs the organic TikTok agency Juiced Media, built the quit-vaping app Puff Count and made the original Wordle iOS app. The thing most worth learning from him is to copy what already goes viral on TikTok, test several formats, and post constantly instead of paying for ads.
if something's already working, if something's already going viral, use those videos as inspiration. It's market research.
The TikTok Organic Strategy That Led to 320K App Downloads
Steven Cravotta: the story and the lessons
1Steven Cravotta and the accidental Wordle app hit
Cravotta started building apps at 14 or 15. An earlier game of his, Grid, reached about 100,000 downloads before Puff Count or Juiced Media.
His forgotten 2015 Wordle app hit 300,000 downloads in one week. The app had earned one to two downloads a week for years. One December week it jumped to 300,000, and he only learned why from a Google search that turned up a New York Times article on Wordle.
He had not updated the app since 2015. He did nothing to cause the surge.
The surge came because the viral Wordle only ran in a browser. Josh Wardle's game had no app, so people searched the app store and found Cravotta's same-named app.
He gave part of the proceeds to an Oakland charity. Through a contact of Wardle's, he donated part of the money to Boost Oakland, an after-school charity.
2How Cravotta grew Puff Count to nearly 300,000 downloads with organic TikTok
Paid ads and paid influencers failed before TikTok worked. For Puff Count, both were too expensive and showed no return. Organic TikTok then worked fast.
His first TikTok flopped. The second got 80,000 views and pushed about 3,000 downloads.
Two years of posting built the audience at $0 in ad spend. The account reached 90,000 followers and over 55 million views, with downloads just under 300,000.
He says these figures are a little outdated. Most of his videos are just him in his apartment trying ideas.
One street interview video took Puff Count to number one in Lifestyle. It reached 5.8 million views in five days and lifted the app to number 30 overall, with no ad spend.
Vaping is sensitive, so nobody agreed to talk for the first hour. He had hired a videographer with a pro camera and mic, and a pair of tourists from Spain finally said yes. The shoot ran two to two and a half hours.
Cheap videos did much of the work. A 30-second trend video took about two minutes of his time, used an already viral sound, and got 1.5 million views and about 8,000 downloads.
A comment reply showing what is in his vape got 8.2 million views and took about 30 minutes. The editing was the hard part.
3Cravotta's method: copy what already works on TikTok
He researches the most liked videos before making anything. He searches a keyword, sets the date to all time, sorts by most liked, and puts the top 20 to 50 videos in a spreadsheet. He notes the hook, storyline and call to action of each.
He does the same for every Juiced Media client and says it works every time. A pickleball search, for example, surfaces one video with 3.8 million likes.
He watches content in his niche before posting his first video. He took the habit from a Gary Vee tweet urging hours of watching your search terms first.
He suggests going through about four pages of results to see what is working.
He tests five or six content formats in month one. Puff Count uses viral trends, a Nic Myths series, public interviews and comment replies. Whichever performs best gets the focus.
Juiced Media tests content for clients, which is how he knew public interviews would work.
He runs a series to keep viewers following. In Nic Myths he states a myth, gives the facts, then promotes the app. Parts one to four push people to follow and watch the rest.
One episode got about 60,000 views. He plans a No Nicotine November series.
He uses hashtags like SEO, not as a viral trigger. A hashtag tells the algorithm who a video is for. A mega viral video can rank first for its hashtag and keep earning views for years.
Videos he posted years ago still rank number one in their hashtags. In his words, "Tik Tok has almost become like this new form of SEO."
He puts a call to action in every video. Even the 30-second trend videos made in his apartment promote the app.
He posts at least daily while warming up an account. Nobody can predict what goes viral, so volume finds the winners. Once momentum builds, he shifts to quality and bigger videos.
He calls TikTok a playground where you can do whatever you want.
He makes heavy topics fun. Quitting vaping is a tough subject, and a simple fun video already had 2.6 million views. He suggests comedy skits for sensitive apps too.
For a family tracker he pitched a skit where a wife tracks her husband to McDonald's just as he walks in wearing gym clothes.
4How Cravotta turns views into users and reuses winners
Most viewers search for the app instead of tapping the bio link. Only 20% click the link in the bio. About 80% close TikTok and search the app store or Google for the app name.
After his last video, Google Search Console traffic skyrocketed. So brand search and the store listing matter as much as the video.
He reposts every TikTok on the other short-form platforms. Vertical video works on Reels, Shorts, Pinterest and Twitter, so one saved video gets posted on five or six other platforms.
He treats organic winners as the best bets for paid ads. Content that earns millions of organic views will likely do well as paid creative, so organic TikTok becomes a cheap testing ground.
He splits creators across accounts, then merges the winners. Each creator has their own style and gets an account. Winning concepts can later move to one main account.
His own choice is one big account, since 90,000 followers build trust. The more you post, the faster you learn.
For apps with weak store search, he leans on outside channels. In an audit of a family location app, he saw searches were mostly branded, like Life360 and find me. He advised TikTok reach plus keywords like family locator, and screenshots that say plainly what the app does.
He would also copy existing viral concepts, such as videos hiding AirTags in kids' shoes, and use the app instead, since a child always has the phone.
He advises delaying account creation so users see the app first. A former SoundCloud employee told him their best account creation rate was 56%. One client raised subscription revenue by showing the paywall before account creation.
Puff Count itself currently makes users create an account.
The TikTok Organic Strategy That Led to 320K App Downloads
Steven Cravotta explains how he grew the Puff Count quit-vaping app with organic TikTok alone, using market research on top videos, repeatable content formats and hashtags. The result is just under 300,000 downloads at $0 ad spend, with one video lifting the app to number one in Lifestyle and number 30 overall.
Cravotta started building apps at 14 or 15. An earlier game of his, Grid, reached about 100,000 downloads before Puff Count or Juiced Media.
His forgotten 2015 Wordle app hit 300,000 downloads in one week. The app had earned one to two downloads a week for years. One December week it jumped to 300,000, and he only learned why from a Google search that turned up a New York Times article on Wordle.
He had not updated the app since 2015. He did nothing to cause the surge.
The surge came because the viral Wordle only ran in a browser. Josh Wardle's game had no app, so people searched the app store and found Cravotta's same-named app.
He gave part of the proceeds to an Oakland charity. Through a contact of Wardle's, he donated part of the money to Boost Oakland, an after-school charity.
Paid ads and paid influencers failed before TikTok worked. For Puff Count, both were too expensive and showed no return. Organic TikTok then worked fast.
His first TikTok flopped. The second got 80,000 views and pushed about 3,000 downloads.
Two years of posting built the audience at $0 in ad spend. The account reached 90,000 followers and over 55 million views, with downloads just under 300,000.
He says these figures are a little outdated. Most of his videos are just him in his apartment trying ideas.
One street interview video took Puff Count to number one in Lifestyle. It reached 5.8 million views in five days and lifted the app to number 30 overall, with no ad spend.
Vaping is sensitive, so nobody agreed to talk for the first hour. He had hired a videographer with a pro camera and mic, and a pair of tourists from Spain finally said yes. The shoot ran two to two and a half hours.
Cheap videos did much of the work. A 30-second trend video took about two minutes of his time, used an already viral sound, and got 1.5 million views and about 8,000 downloads.
A comment reply showing what is in his vape got 8.2 million views and took about 30 minutes. The editing was the hard part.
He researches the most liked videos before making anything. He searches a keyword, sets the date to all time, sorts by most liked, and puts the top 20 to 50 videos in a spreadsheet. He notes the hook, storyline and call to action of each.
He does the same for every Juiced Media client and says it works every time. A pickleball search, for example, surfaces one video with 3.8 million likes.
He watches content in his niche before posting his first video. He took the habit from a Gary Vee tweet urging hours of watching your search terms first.
He suggests going through about four pages of results to see what is working.
He tests five or six content formats in month one. Puff Count uses viral trends, a Nic Myths series, public interviews and comment replies. Whichever performs best gets the focus.
Juiced Media tests content for clients, which is how he knew public interviews would work.
He runs a series to keep viewers following. In Nic Myths he states a myth, gives the facts, then promotes the app. Parts one to four push people to follow and watch the rest.
One episode got about 60,000 views. He plans a No Nicotine November series.
He uses hashtags like SEO, not as a viral trigger. A hashtag tells the algorithm who a video is for. A mega viral video can rank first for its hashtag and keep earning views for years.
Videos he posted years ago still rank number one in their hashtags. In his words, "Tik Tok has almost become like this new form of SEO."
He puts a call to action in every video. Even the 30-second trend videos made in his apartment promote the app.
He posts at least daily while warming up an account. Nobody can predict what goes viral, so volume finds the winners. Once momentum builds, he shifts to quality and bigger videos.
He calls TikTok a playground where you can do whatever you want.
He makes heavy topics fun. Quitting vaping is a tough subject, and a simple fun video already had 2.6 million views. He suggests comedy skits for sensitive apps too.
For a family tracker he pitched a skit where a wife tracks her husband to McDonald's just as he walks in wearing gym clothes.
Most viewers search for the app instead of tapping the bio link. Only 20% click the link in the bio. About 80% close TikTok and search the app store or Google for the app name.
After his last video, Google Search Console traffic skyrocketed. So brand search and the store listing matter as much as the video.
He reposts every TikTok on the other short-form platforms. Vertical video works on Reels, Shorts, Pinterest and Twitter, so one saved video gets posted on five or six other platforms.
He treats organic winners as the best bets for paid ads. Content that earns millions of organic views will likely do well as paid creative, so organic TikTok becomes a cheap testing ground.
He splits creators across accounts, then merges the winners. Each creator has their own style and gets an account. Winning concepts can later move to one main account.
His own choice is one big account, since 90,000 followers build trust. The more you post, the faster you learn.
For apps with weak store search, he leans on outside channels. In an audit of a family location app, he saw searches were mostly branded, like Life360 and find me. He advised TikTok reach plus keywords like family locator, and screenshots that say plainly what the app does.
He would also copy existing viral concepts, such as videos hiding AirTags in kids' shoes, and use the app instead, since a child always has the phone.
He advises delaying account creation so users see the app first. A former SoundCloud employee told him their best account creation rate was 56%. One client raised subscription revenue by showing the paywall before account creation.
Puff Count itself currently makes users create an account.