I. What Is PicPet
PicPet (O3O Labs, Taiwan) — two people co-raise a virtual pet by feeding it photos. Stop sharing photos, the pet dies. BeReal mechanic meets Tamagotchi emotional stake. Launched August 2025, spread organically in Taiwan via Threads.1
DAU (Taiwan)
400K
self-reported
Downloads
623K+
App Store
HK Launch
Feb 11
6 days ago
II. Recommended Posts — Ready to Use
These posts are modeled on actual HK viral Threads posts (Section III), not Taiwan patterns. Each explicitly references the HK benchmark it mimics. Replace [code] with your PicPet pet invite code.
Post A — Confession + Challenge PRIMARY
Modeled on @kyy05_25 (69,700 likes) — boyfriend challenge mechanic + @kekeilubu013_ (12,500 likes) — raw couples confession format
Ready-to-Post · HK Cantonese · Confession + Challenge
我男朋友成日已讀唔回。
見過佢打機打到天光。
見過佢放假瞓到下午三點。
見過佢隔三日先沖涼。
但叫佢影張相俾我就好似要佢條命。
好。我搵到方法。
下載咗 PicPet。兩個人一齊養隻寵物。
隻寵物要食相先活到。唔餵就死 💀
而家佢日日主動send相俾我。
早餐、lunch、返工條路、廁所selfie都有。
交易係咁嘅：
每日佢漏餵一次 = 佢請食飯 🍽️
隻叫「年糕」。仲未死過。
佢請咗我食三次。
寵物序號：[code]
姐妹們。你哋試。實測有效 ✅
Why this works: Opens with the HK viral "list of gross/real things about my boyfriend" format (@kekeilubu013_, 12.5K likes). Adds a challenge mechanic with real stakes ("漏餵一次 = 請食飯") mirroring @kyy05_25's "每個like = 10蚊" format (69.7K likes). Self-deprecating, raw, no marketing polish. Ends with a pet code for viral loop.
Post B — Raw Discovery Story ALTERNATIVE
Modeled on @justeunme (17,000 likes, 8,900 shares) — couples anecdote with punchline format
Ready-to-Post · HK Cantonese · Anecdote
同男朋友試咗個台灣嘅app。
佢話佢好忙冇時間理我。好。
PicPet。兩個人養隻寵物。要影相餵。
第一日佢影咗碗公仔麵俾我睇。
第二日影佢對臭襪。
第三日影佢老細個後腦。
我問佢點解影啲咁嘅嘢。
佢話：「隻嘢肚餓我驚佢死呀嘛」
三日。由已讀唔回變成日日send相。
台灣400K人玩緊。香港啱啱先開通 🇭🇰
我哋隻叫「叉燒」。佢依家level 4。
寵物序號：[code]
📱 PicPet — App Store
Why this works: Mirrors the @justeunme format — short couples anecdote building to an absurd punchline ("佢老細個後腦"), then the reveal. The humor comes from the boyfriend's ridiculous photos, not from the app pitch. The "已讀唔回" → "日日send相" arc is the emotional payoff. Highest share-to-like ratio pattern (0.52x) — people send this to their partner as a hint.
Usage notes
- Replace
[code] with your actual PicPet pet invite code
- Attach a screenshot of your pet — TW posts with images get 2–4x more shares
- Post between 8pm–11pm HKT (peak Threads engagement)
- Post A is the stronger bet — challenge mechanics drove the #1 HK Threads post this month (69.7K likes)
- The key insight from HK Threads: don’t recommend the app. Share your story of using it. Product recs get <200 likes. Personal confessions get 10K+.
III. What Actually Goes Viral on HK Threads
Before writing a single word, we scraped the highest-performing HK Threads posts to understand what actually gets engagement. These are the benchmarks our posts are modeled on — not Taiwan patterns, not assumptions.
The Benchmarks
“呢種好事都終於係輪到我🤓 我男朋友話呢條threads有幾多個like 就俾幾多錢我 1個like 10蚊港紙 截至到14號 交俾大家了”
2Pattern: Challenge mechanic · boyfriend as prop · audience participation · Feb 7, 2026
“男朋友朝早起身扯旗 但因為我嚟m 所以叫佢自己打飛機 佢唔想 但佢越嚟越硬 我就叫佢同我一齊背主禱文 念完佢就瞬間軟了 我哋兩個笑到傻”
3Pattern: Raw couples anecdote · absurd punchline · share-to-like ratio 0.52x · highest share rate
“我好佩服我男朋友 見過我喊 見過我嘔 見過我痾屎 見過我放屁 見過我爆瘡 見過我素顏 見過我變肥 見過我瞓到死死下 見過我壓力大到crash out 佢到依家都仲鍾意我真係好嘢”
4Pattern: Accumulation list · gross/real details · earnest punchline · zero English except “crash out”
“我自行宣布 第一屆「實用可一次買斷app」大賽 正式開始！ 選手如下：”
5Pattern: Crowd-sourced list challenge · app discovery format · 8,600 replies = massive engagement loop
“到我分享 月薪 🐼🙌🏻 香港全職做vtuber第3年！ 試過最高峰月入10萬 冇加入任何企業...為咗搞3D演唱會，自己學埋coding”
6Pattern: Personal disclosure · behind-the-scenes honesty · HK-specific · local creator economy
The Patterns
| Pattern | What It Is | Engagement | Our Posts |
|---|
| Challenge mechanic | Boyfriend/audience stakes tied to post performance | 69.7K likes | Post A: “漏餵一次 = 請食飯” |
| Accumulation list | Pile up relatable/gross details, earnest punchline | 12.5K likes | Post A: “見過佢...” opening |
| Absurd anecdote | Setup → escalation → ridiculous punchline | 17K likes | Post B: boyfriend’s photo progression |
| Crowd-sourced list | “I declare the first annual [X]” — invite replies | 104K likes | Future post: “PicPet寵物名大賽” |
| Personal disclosure | Behind-the-scenes honesty about real life | 2.2K likes | Both posts: real details, no marketing voice |
What does NOT work on HK Threads: Product recommendations in review format (<200 likes). Polished marketing copy. TW Mandarin content (immediately pattern-matched as “not from here”). Cute/exclamatory tone (“超好用！！”). Everything that looks like an ad.
IV. The Only Proven HK Threads Marketing Playbook
In March 2025, the Hong Kong film Peg O’ My Heart (贖夢) ran the only well-documented viral Threads campaign in HK. The results are the best available guide for what works.7
Threads Share
53.7%
of all film buzz
Peak Spike
1,800
mentions/day
Strategy
In-Character
not product push
Top Format
Voice Notes
beat text + images
What They Did
Actor-director Nick Cheung created a Threads account as his fictional character (蔡辛強, a sleep-deprived taxi driver). He posted in-character, responded to fans in-character, and used voice notes extensively. He never once said “go watch my movie.”
When someone photographed him “napping on the MTR,” his in-character response (“已經瞓唔到覺，仲要搭車瞓覺都畀人偷拍！”) generated 1,800 mentions the next day — the campaign’s first viral spike.
What This Means for PicPet
Don’t Do This
- Post “PicPet is a great app you should try”
- List features (photo-sharing, pet leveling, etc.)
- Use marketing language (“revolutionary,” “must-try”)
- Post as a brand or marketer
Do This Instead
- Post as a USER sharing a personal story
- Talk about your boyfriend, not the app
- Let the app be the incidental discovery
- Use voice notes for follow-up engagement
The Peg O’ My Heart rule: The product is the backdrop, not the subject. Nick Cheung’s character drove 53.7% of all buzz without once saying “go watch this movie.” PicPet posts should be about the couple’s story — the app is discovered incidentally by the reader.
V. PicPet’s Viral Mechanics (Taiwan Context)
While TW patterns don’t directly transfer to HK voice, PicPet’s product mechanics are universal. These are the six elements that make the app inherently viral:8
| # | Element | How It Drives Growth |
|---|
| 1 | Tamagotchi loop | Emotional stake — your shared pet depends on you |
| 2 | Photo = minimum effort | Snap anything — lowest possible friction to participate |
| 3 | Friend-gated activation | Must invite someone to start — built-in viral loop |
| 4 | Familiar emoji pets | No design learning curve, instant recognition |
| 5 | Random pet gacha | Surprise at which pet you get — shareable moment |
| 6 | Leveling system | Progression = retention + bragging rights |
Top TW post: @peter_0316_ — “今天中午可能所有的高中生都在玩這個吧” — 11,700 likes, 5,000 shares.9 The pet invite code in that post is the conversion mechanism — readers tap it and start playing immediately. Both recommended posts above include a code placeholder for this reason.
VI. The HK Gap
PicPet became available in Hong Kong on February 11, 2026.10 As of this research (Feb 17), we found one HK-specific Threads post:
“picpet 在香港可以玩了！”
Feb 11, 2026 · only HK-specific PicPet post found across all of Threads
For context: the top TW PicPet post has 11,700 likes. The HK one has 20. The gap between TW saturation and HK awareness is enormous. CNY week (Feb 17–23) is peak couples content time.
VII. HK Threads — Platform Data
From a 212-respondent survey of HK Threads users and the Marketing-Interactive analysis:1112
Prefer Text
86.8%
over images/video
Use For
84.9%
entertainment & gossip
Post Frequency
62.7%
<1x per week
Daily Logins
70.3%
3+ times/day
What this means: Most HK users are consumers, not creators. They scroll 3+ times daily looking for entertainment. Only a minority post. Content that gets engagement must be entertaining enough to make the 62.7% silent majority stop scrolling — and our benchmarks show that means raw, personal, funny content, not product reviews.
Additional signals: 68.4% prefer following personal accounts over brands. 51.4% feel Threads reflects HK local culture — meaning there’s a gap for content that feels authentically Hong Kong. Some users now use Threads as a substitute for LIHKG (HK’s major forum).
VIII. Launch Playbook
Step 1: Post as a User, Not a Marketer (Day 1–3)
The Peg O’ My Heart lesson: the product is discovered incidentally. Publish Post A from a personal account — no brand handles, no app tags. Share a real experience with PicPet. Attach a pet screenshot. Include the invite code at the end, not the beginning.
Step 2: Challenge Format Seeding (Day 2–5)
The @eting15 post (104K likes) proves crowd-sourced challenges are the highest-engagement format on HK Threads. A natural evolution: “PicPet寵物名大賽 — share你隻寵物叫咩名” invites replies, each reply includes a pet code, and the viral loop compounds.
Step 3: Pet Code Chain Reaction (Day 3–7)
In Taiwan, users post pet codes publicly asking strangers to co-raise pets. This mechanic transfers — the code is the conversion CTA. Once initial posts gain traction, encourage early adopters to share codes on Threads. This is the self-sustaining loop.
Timing. CNY week = couples content peaks, people have time to try new apps, sharing photos of holiday activities is natural PicPet content. But also: 62.7% of users post less than once a week. Don’t expect instant flood. The first well-placed post during CNY has the best chance of catching the algorithm. One great post > five mediocre ones.
Bottom Line
PicPet’s product mechanics are inherently viral (pet codes, tamagotchi guilt, friend-gated activation). Hong Kong launched 6 days ago with zero localized content. The window is open.
But the path to HK viral is not the same as Taiwan. HK Threads rewards raw confessions and challenge mechanics — not cute endorsements or polished app reviews. The posts above are modeled on what actually works: the 69.7K-like challenge format, the 17K-like absurd anecdote, the 12.5K-like accumulation list. Post as a user. Tell your story. Let the app be the surprise.
Sources
[1]
PicPet — App Store (Taiwan) — O3O Labs Inc., 623K+ downloads, 4.8 rating, launched Aug 4 2025[2]
@kyy05_25 — “我男朋友話呢條threads有幾多個like就俾幾多錢我” — 69,700 likes, 702 replies. HK boyfriend challenge post. Feb 7 2026[3]
@justeunme — HK couples humor (raw anecdote) — 17,000 likes, 8,900 shares. Highest share-to-like ratio in benchmark set. Apr 23 2025[4]
@kekeilubu013_ — “我好佩服我男朋友” accumulation list — 12,500 likes, 6,700 shares. List-of-things format. Nov 4 2025[5]
@eting15 — “第一屆 實用可一次買斷app 大賽” — 104,000 likes, 41,100 shares. Highest-engagement app-related post on Threads. Nov 25 2025[6]
@mya.desu — HK VTuber salary disclosure — 2,200 likes, 152 replies. Personal disclosure format. Aug 25 2024[7]
Marketing-Interactive — Peg O’ My Heart viral Threads case study — uMaxData analysis: 53.7% of buzz from Threads, 85 posts, voice notes = highest engagement format. Mar–Apr 2025[8]
@miblue119 — PicPet viral mechanics analysis — 56 likes. Identifies 6 interlocking viral elements. Nov 5 2025[9]
@peter_0316_ — “今天中午可能所有的高中生都在玩這個吧” — 11,700 likes, 5,000 shares. Top PicPet post in TW. Nov 28 2025[10]
@scnrynpw — “picpet 在香港可以玩了！” — 20 likes. Only HK-specific PicPet post found. Feb 11 2026[11]
Marketing-Interactive — HK Threads survey (212 respondents) — 86.8% prefer text, 84.9% for entertainment, 62.7% post <1x/week. By Rudi Leung, Hungry Digital[12]
RFI Asia — Why Threads Finally Matters for Marketers — 81% daily login rate HK, 71% check multiple times/day. 190M global users[13]
@dou_o6l6 — TW couples endorsement — 4,100 likes, 542 shares. TW archetype (not HK voice). Dec 3 2025[14]
@_____8.14 — TW personal discovery story — 2,800 likes, 4,200 shares. TW archetype. Jun 28 2025[15]
PicPet — Spark App Intelligence — download and rating data[16]
Business Next (TW) — PicPet coverage — 400K DAU Taiwan, referenced by Jovy Yu