⚠️ Use Product Hunt with caution for 2C products
This skill defaults to a dev / maker / B2B audience.2C consumer goods / education / apps / games: The Product Hunt audience is makers/tech people, a mismatch with ordinary consumers, with extremely low ROI—launch at most once to get a badge. Please redirect your effort to vertical communities (Reddit / Xiaohongshu / Zhihu / Naver 카페) + short video (TikTok / Douyin / YouTube) + vertical KOLs, and use "free product as a hook" to drive conversion. For the complete 2C channel database (country-by-country MAU + public sources) + the three-way Launch branch, see → gingiris-seo-geo/references/2c-adaptation.md(and gingiris-launch)。
gr-ph-launch — Product Hunt launch
In-depth reference
For detailed methodology and case library, see:
references/framework.md— 14-day timeline + 5 major sectionsreferences/hunter-playbook.md— hunter selection matrixreferences/maker-comment.md— 6 maker comment templates
Note: On first use, Claude should read
references/framework.mdbefore starting.
14-Day Timeline (Simplified Version)
T-14: Preparation
- Product page ready (landing page, pricing, FAQ)
- Already have ≥ 100 early-bird users + email list
- Choose 1 hunter (priority: active > follower count)
- Prepare 3 main gallery images + 1 30-second demo video
T-7 to T-2: Warm-up
- Post 1 "we're launching, AMA" in the PH discussion section
- Tease on Twitter / LinkedIn / Reddit (do not post the launch page link, to avoid stealing traffic)
- Say hello to users via newsletter
- Confirm hunter time zone + launch time (PST 00:01)
T-1: Final check
- Lock final versions of all assets
- Prepare the first comment draft (within 500 words)
- Prepare 10 possible FAQ reply templates
- Contact 10-20 core supporters to remind them about tomorrow morning
T-0: Launch day
- 00:01 PST hunter launches
- 00:15 maker (you) posts the first comment
- All day: reply to a batch of comments every 30 minutes; do not bot upvotes
- 14:00 PST: peak time, push the second wave (contacts + community)
- 23:00 PST: post a thank-you note as a summary
T+1 to T+7: momentum
- Reply to all comments (including negative ones)
- Turn the PH featured screenshot + data into a case study
- If you reach daily #1 → write a "how we won" blog post (
gr-blog-post) - Import all upvoter emails into the user pool
Hunter Selection Matrix
| Priority | Criteria |
|---|---|
| P0 | Has had a successful hunt in the last 30 days and the product field is similar |
| P1 | Follower 1k+ and active (posts weekly) |
| P2 | Has a real social relationship with you (have chatted, mutual friends) |
| ❌ | Follower > 50k but has not hunted for 6 months |
Search hunter:https://www.producthunt.com/@<category>-hunters
6 Maker Comment Templates
(See details in references/maker-comment.md)
- Founder story: why build it → your pain → solution
- Before / After: before vs after having your product, sharp contrast
- Key stats hook: Start with an explosive data point
- Honest limitation: Admit the product isn't perfect yet, which makes it feel authentic
- Community thanks: Thank the users who participated in testing (name @ them)
- Tech stack pride: Talk about the tech stack (suited to developer-facing products)
Which to choose: depends on the target audience. To dev → 6; to general → 1 or 3.
Anti-landmine checklist
- ❌ Don't still be changing the product on launch day (flaky)
- ❌ Don't let family and friends spam upvotes (PH shadowbans fast)
- ❌ Don't leave external links in the first comment (triggers moderation downranking)
- ❌ Don't forget mobile screenshots (40% of traffic comes from phones)
- ❌ Don't force yourself to hold daily #1; #2–#3 is also a success (and gets you less flak)
Upvote pacing red lines (2026 tested updates)
- Total ≤100 votes/hour, single channel ≤30 votes/hour——PH clears suspicious votes about every 2 hours, so spikes will be stripped
- Supporters should vote using **mobile + cellular data (4G)**, not all crowding the same office Wi-Fi
- Badge exposure tiers: weekly ranking ≈ 7x daily ranking, monthly ranking ≈ 30x——Decide which ranking to push for first, then set the pace; for the monthly ranking, choose the Tuesday of the second/third week of each month
Go / no-go: budget tier benchmarks
| Budget | Realistic output |
|---|---|
| ~$1K | A few hundred signups (skip PH, go directly to KOC + vertical communities) |
| $5K | About 1,800 votes in 24h is achievable——real case: after hitting daily #1, PH later cleared votes and took it back. The pacing red lines must be followed |
| $10–20K | ~1 million impressions + a few thousand signups |
If the budget isn't high, don't grind on PH: With the same budget, giving exposure to ~10 Twitter KOLs/KOCs yields higher impressions and signups——PH exposure→upvote conversion is only ~1%, while Reddit open-source promotion exposure→star conversion is 5–8%.
Launch-day traffic reverse planning method
First set the total traffic pie → break down hard numbers per channel → attribute via UTM short links per channel. Actual test (an AI Agent product, 2025-12): 500K traffic = 400K Twitter (40 influencer quotes) + 20K Reddit + 20K YouTube paid promotion + 20K groups + the rest → about 800 signups, CAC $6.25。
Advanced pacing:two-day structure(half the influencers drive to the official site on launch day, the other half drive to PH on ranking day);concentrated KOL burst(50–100 KOLs agree to post densely within 24–48h).
Cascading recommendations
- Launch day + 24h →
gr-seo-patrolmonitor rankings for keywords like "product hunt" - After hitting daily #1 →
gr-blog-postwrite a "how we won" post and set the canonical to the main one - If there is a lot of skepticism in the comments →
gr-user-interviewstart formal interviews
API dependencies
| Service | Env var |
|---|---|
| Product Hunt (public page scraping) | None (use web-access skill) |
| GitHub PAT (blog linkage) | GITHUB_TOKEN |