Push-up yokai slayer: a Japanese developer took an overseas "push-up battle" concept, gave it a Japanese folklore skin, and a TikTok pushed it to #6 on the Fitness chart

This summer, a wave of "phone camera counts push-ups + online 1v1 battles" apps appeared in Europe and the US, but nobody in Japan built one seriously. In Oct 25, Japanese indie dev Masaya Fukuyama launched 「腕立て妖怪退治」 (Pushup Youkai Taiji). The mechanic was the same, but the wrapper was "bad habits are youkai haunting you, do push-ups to slash them," and it shipped in Japanese only. A TikTok with 100k views pushed it to #6 on the Japan Health & Fitness free chart, and then its rank kept sliding. Below: what it actually did right, why the buzz didn't hold, and what you can take from it.

A 28-second video

On Oct 26 2026, just after 1 p.m. Japan time, the second day after launch, the official TikTok account @udetate_youkai posted a 28-second video with one line as its title: 「あなたはこれに勝てる？」 (Can you beat this?). The footage was a 30-second push-up battle recording.

The video reached 11-06 views, 105,321 likes, and 1,123 shares (TikHub, 60 11-06). The most-liked comments tell the story: 「こんなんで腕立てカウントされるんだ」 (Does it really count push-ups like this?), 「なんて言うゲーム？」 (What's this game called?), 「なんて調べればでますか？」 (What do I search for to find it?), and one saying 「これズルできちゃうよな」 (You can probably cheat with this). The first reaction was "I want to try it," the second was "go search for it in the store."

The App Store Japan storefront moved too: 10-28 it was already at #6 on the Health & Fitness free chart, and its rating count went from 10-28 9 ratings to 11-06 118 ratings (App Store Japan chart and rating counts). But this account never posted a second big hit after that.

What it is

You stand the phone upright on the floor in front of you. The front camera tracks your head as it goes up and down, and every push-up slashes a monster. No equipment, nothing to wear.

The store page's first line is 「悪い習慣は、妖怪のしわざ」 — bad habits are the work of yōkai. Each night you meet a new bad-habit yōkai: 二度寝狸 (a tanuki that makes you go back to sleep), 夜更かし入道 (a night-owl monk), あとでやる河童 (the "I'll do it later" kappa), スマホろくろ首 (a phone-scrolling long-neck yōkai) and so on. Each chapter's Boss is called 「三日坊主」「七日坊主」「三十日坊主」「百日坊主」「千日坊主」 — in Japanese, 三日坊主 means a three-day wonder, someone who quits after three days. The whole game is about one thing: this time, don't quit on day three.

Around that story, it reskins the usual habit-app playbook: XP and levels, special moves, and you can pick a training monk, a shrine maiden, or an onmyōji as your exorcist, with five appearance changes tied to level. A streak of 7 days gives you a 「身代わり地蔵」 (Migawari Jizo, a stand-in guardian statue) that protects your streak on any day you skip training. There's also a venting mode called 「ムカつき提灯」 (Irritation Lantern): write down what annoyed you today, like 「締切」 (deadline), 「月曜日」 (Monday), 「満員電車」 (packed train). It turns into a lantern youkai, you do push-ups to knock it down, and the result image can be shared on social media.

What actually drove its spread was online battles: 30-second push-up 1v1 matches, with nationwide rankings. 11-05's 3.3 version added monthly "seasons"; if you play 10 matches in a season, you earn 「位の札」 (rank cards), and the top ten with 30+ matches get the titles 「天下一」 (Tenka Ichi), 「三傑」 (Sanketsu), and 「十傑」 (Jikketsu) (App Store release notes, 11-05). In 11 days after launch, the app went from version 1.2 to 3.3.

It didn't invent this idea.

Camera rep counting, boss fights, 1v1, leaderboards: all of this already exists overseas, and even the TikTok playbook is ready-made. RepChamp's account posts battle clips every day with "Can you beat this?" captions, and 「腕立て妖怪退治」's 「あなたはこれに勝てる？」 is nearly a direct translation.

It won Japan. Peer apps have rating counts in the 0 to 39 range in the Japan store, and it hit 11 ratings in 120 days, making it the top app in this small Japan niche. The overseas apps are either English-only or machine-translated multilingual versions, and none tell a story Japanese players instantly understand. One 5-star review reads 「なんで今までこんな神アプリが無かったのか」 (Why didn't a godly app like this exist before, 10-28).

Where users come from

Almost entirely from the official TikTok account. By 11-06, @udetate_youkai had only 62 followers, 14 videos, and 1,273 likes in total, 1,123 of which came from that one hit (TikHub, 11-06). After that it posted almost daily, all battle recordings with provocative titles: 「60歳の父親が強すぎる」 (My 60-year-old dad is too strong), 「ハロウィンにマッチした五条悟強すぎる」 (The Gojo Satoru I matched with on Halloween is too strong), 「アンチと腕立て対決してみた」 (I did a push-up duel with a hater), 「視聴者と直接対決の結果」 (Results of a direct showdown with viewers). Views went from 459 to 6,173, and no second big video ever came.

Its chart rank followed this curve: 10-28 #6, 10-30 #19, 11-01 #58, 11-04 #55, 11-06 #99 (App Store Japan Health & Fitness free chart).

A little organic user sharing. Regular users posted their results under the #腕立て妖怪退治 hashtag. One account posted 7 times, each video getting three to five hundred views, with a bio reading 「腕立て5勝2敗なう」 (Push-ups: 5 wins, 2 losses, and counting). Battle results are naturally something people want to share, but the volume is still very small.

Store search. The subtitle packs in two search terms, 「筋トレ習慣化」 (building a workout habit) and 「腕立て伏せカウンター」 (push-up counter). The TikTok bio also tells people to search for the name in the App Store.

Two inferences. First, its rank fell to #99 while it still gained 6 to 24 new ratings per day. That suggests a rating prompt in the app, or a return-visit moment like seasons triggering ratings (inference). Second, the comments on the 100k-view video number only 11, which is low for that view count, so paid promotion is possible (unverified).

The developer isn't new to this. Under the same account there are 12 apps: DoggyWalk, a dog-walking tracker (183 ratings), Loving, a couples widget (83 ratings), BoxScore+ for basketball scorekeeping, and 「スクワットロック」 (Squat Lock), which blocks social apps until you do squats. They're all small utilities. This is the first time he's turned exercise into a game (iTunes, 11-06).

Where the money comes from

The first 7 monsters are free. Paying unlocks 4 more chapters with 28 monsters, catch-up check-ins, training mode, squats, and history charts. Annual plan ¥3,600 (with a 2-week free trial), monthly ¥780, one-time purchase ¥5,800. Costumes are sold separately for ¥160 to ¥480 (official site).

It has never reached the grossing charts, its revenue isn't public, and it's most likely small (estimate).

Its weak spots

• Counting isn't accurate. The bad reviews cluster here: 「反応ゴミすぎる / ちゃんと腕立てしてるのに反応しない」 (The response is garbage, it doesn't react even though I'm doing push-ups properly, 10-27). In battles, this is a fairness problem. One 3-star review says 「倍以上腕立てしてるのに反応の悪さで負ける」 (I did more than double the push-ups and still lost because it responded poorly).
• Not enough players, so matchmaking fails. The same review says 「人が少なく対戦マッチがしない事が大半」 (Most of the time there are too few people and no match happens). Battles are its sharing hook, and they're also the part that depends most on user numbers.
• It didn't keep up after one hit. It never charted outside Japan, and its Japan rank dropped from #6 to #99 in ten days.

If it were you

C. What's worth watching in this app is the approach, not the business. It takes a category already validated on TikTok overseas, wraps it in a local cultural skin that locals instantly understand, and supports only one language. That's enough to reach #1 in a small niche in one country. But the push-up battle category itself hasn't proven it can make money: the biggest overseas peer in the US store has barely over a thousand ratings, and this app never reached the grossing charts itself. Its rank dropped off within ten days.

Who it suits: one or two people who can build camera-based motion recognition (iOS Vision body pose is good enough), and who know a market with strong local culture where overseas apps aren't localized, such as Korea, Taiwan, or a Southeast Asian country. They need to write "bad habit monsters" that make locals smile on sight.

Who it doesn't suit: anyone aiming to go head-on at the English market (RepChamp and Pushup RPG are already crowded there); anyone who wants it as a main business relying on subscriptions to pay the bills (there are no revenue signals at all); anyone who can't get recognition accuracy right (inaccurate counting in battles means an automatic loss, and that's where one-star reviews come from).

How to build it: don't copy the Japan version, copy the method. Pick a market, swap the "three-day willpower" pain point for local phrasing and local monsters. Keep the core loop free, one new monster each night, a catch-up check-in every 7 days, and sell subscriptions for later chapters and for protecting streaks. For battles, start with "race a recorded real person's score" instead of real-time matchmaking. When there are few users, real-time 1v1 just makes people wait.

How to get users: as it and RepChamp did, collect battle recordings before launch, post the strongest one on launch day, and use a provocative question as the title ("Can you beat this?"). After that, post every day, rotate opponents and personas (dad, night owl, haters), and let users share their results image in one tap.

Deadline: there's no urgent window. Anyone can do this reskin approach. The advantage of being first in one country probably lasts only a few months (inference).

How this call could be wrong: ① If after December it still holds around #50 or better in the Japan Health & Fitness free chart, or makes the grossing chart, then this isn't just a one-time hit. Seasons are keeping people around, the business is tougher than it looks, and the grade should go up. ② If an overseas push-up battle app reaches the top 100 of the US Health & Fitness grossing chart, the category itself makes money, and a reskin for other countries would be a B or even A. ③ The other way around: if it drops out of the top 200 of the Japan Health & Fitness chart before December and the updates stop, then this was just a single TikTok.

---

Data and sources: App Store Japan Health & Fitness free chart, rating counts, and reviews (10-28 to 11-06); iTunes lookup (Japan and US stores, 11-06: rating counts, versions, release notes, developer's app list); store page description; official site fukuwebstudio.jp/mikka; TikTok @udetate_youkai videos and comments, @repchampapp, the #腕立て妖怪退治 hashtag (TikHub, 11-06); competitor sites repchamp.vercel.app, pushupboss.app/ja, pushup.quest. Wherever "estimate," "inference," or "unverified" appears, that's what it is.