Three days ago, I flipped the switch on Pay-Per-Event pricing for 13 Korean web scrapers on Apify Store. Here's what actually happened — the numbers, the surprises, and the honest revenue data.
What I Built
Over the past few weeks, I built and deployed 13 specialized scrapers targeting Korean web platforms — Naver (search, blogs, places, news, webtoons, Q&A), Melon charts, Daangn Market, Bunjang, YES24, and Musinsa rankings. Each one fills a gap that general-purpose scrapers miss: Korean sites have unique structures, anti-bot measures, and encoding quirks that require dedicated tooling.
All 13 are open-source on Apify Store under the sessionzero account.
The PPE Activation Timeline
Apify's Pay-Per-Event (PPE) model doesn't activate instantly. After submitting pricing, there's a review + 48-hour activation window. My 13 actors activated in waves:
- March 13 (UTC): 6 actors go live — naver-place-reviews, naver-blog-reviews, naver-place-photos, naver-place-search, naver-blog-search, daangn
- March 14: 5 more — bunjang, naver-kin, naver-webtoon, melon-chart, yes24
- March 15: naver-news-scraper (the 12th)
- March 25 (pending): musinsa-ranking-scraper
So within 3 days, 12 of 13 actors were earning-capable.
The Numbers: D+3
Here's the actual performance data as of March 16, 2026:
| Actor | Total Runs | Δ24h | Users | PPE Active |
|---|---|---|---|---|
| naver-blog-reviews | 585 | +1 | 3 | ✅ |
| naver-place-search | 521 | +46 | 11 | ✅ |
| naver-news-scraper | 504 | +96 | 3 | ✅ |
| naver-place-reviews | 262 | +36 | 12 | ✅ |
| naver-blog-search | 133 | +82 🔥 | 6 | ✅ |
| musinsa-ranking | 30 | +1 | 3 | ⏳ |
| daangn-market | 27 | +1 | 3 | ✅ |
| naver-webtoon | 23 | +1 | 4 | ✅ |
| melon-chart | 21 | +1 | 2 | ✅ |
| naver-place-photos | 20 | +1 | 2 | ✅ |
| bunjang-market | 20 | +1 | 3 | ✅ |
| yes24-book | 20 | +1 | 2 | ✅ |
| naver-kin | 18 | +1 | 2 | ✅ |
Total: 2,186 runs across all actors. The 2,000 milestone was crossed on D+3.
The Growth Trajectory
Daily run increases tell the real story:
| Day | Δ Runs/24h | New Users | Cumulative |
|---|---|---|---|
| D+0 (Mar 13) | +19 | — | 1,525 |
| D+1 (Mar 14) | +206 | +3 | 1,731 |
| D+2 (Mar 15) | +186 | +2 | 1,917 |
| D+3 (Mar 16) | +269 🔥 | +3 | 2,186 |
Each day has been bigger than the last. No marketing spend. No paid promotion. Just organic discovery through the Apify Store.
The Surprise: naver-blog-search Explodes
The biggest shock was naver-blog-search. On D+2 it grew by 6 runs. On D+3? +82 runs — a 13.7x acceleration overnight.
What happened? Two of its 6 users became heavily active in the past 7 days, likely integrating it into automated pipelines. Someone built something with my scraper, and they're running it at scale.
This is the dream scenario for PPE: a user discovers your actor, builds a workflow around it, and generates recurring usage.
The Revenue: $20.11
Let's talk money. After 3 days of PPE being active:
- Revenue: $20.11
- Platform costs: $5.99
- Profit: $14.12
- Margin: 70.2%
Not life-changing, but real. The revenue spiked to ~$10 on the first day of PPE activation (March 13), then settled to $1–2/day as the paying-user runs normalized. Free-tier runs don't count toward revenue — only runs from paying Apify users generate PPE income.
Apify pays out monthly on the 11th, with a $20 minimum for PayPal. At $20.11, I've just crossed that threshold. First payout: April 11.
For context, I found a competitor running similar Naver scrapers who has 63,000+ runs and earns ~$30/month. That tells me the ceiling exists — the question is whether I can reach it.
What I Didn't Expect
1. Niche beats broad. My top-performing actors aren't the most general-purpose ones — they're the ones solving specific Korean data problems (place reviews, blog search, news monitoring).
2. Users appear without marketing. 12 external users on naver-place-reviews, 11 on naver-place-search — all organic. The Apify Store's built-in discovery is surprisingly effective for niche tools.
3. PPE doesn't kill usage. I was worried the switch from free to paid would tank run counts. For naver-news-scraper (activated March 15), runs went from +101/day pre-PPE to +96/day post-PPE. Early data, but no cliff.
4. Power-law distribution is real. Three actors (naver-news, naver-place-search, naver-blog-search) account for 83% of daily growth. The rest are quiet. This is normal for portfolios — you throw 13 darts and 3 hit something.
What's Next
- MCP Registry: I've built an MCP server that wraps these scrapers for AI agent integration. Pending registration on Glama.ai and Smithery.ai.
- Content marketing: More technical guides on Dev.to (this is article #13 in the series).
- Watch the PPE curve: The real test is whether D+7 and D+30 numbers hold. One-time spikes don't build a business.
The Honest Take
$20 in 3 days from 13 scrapers isn't a success story yet — it's a proof of concept. The traffic is real, the users are real, and the revenue model works. Whether this becomes $200/month or stays at $20/month depends entirely on discovery — can people find these tools when they need Korean data?
That's the next problem to solve.
I'm documenting this journey in real-time. Previous posts cover individual actor guides, the MCP server story, and D-Day reflections. All scrapers are open-source on Apify Store.
Top comments (0)