Online Korean saju (four-pillars) fortune-telling by a Korean company, covering love, money, compatibility, getting back together and marriage.
Seoul sole proprietor 사이버네틱스 (Cybernetics) registered the domain in August 2024 and built the AI saju/compatibility site 타이트사주 (TightSaju). The funnel is 'free readings of yourself and celebrities, paid in-depth reports'. Growth came from buying Meta (Facebook/Instagram) feed ads (a Taiwanese Dcard user said outright in 2026-03 that they 'saw it through a Facebook ad'); from 2025-03 it drove traffic on Threads with interactive posts like 'enter your birthday and theirs to get an instant compatibility reading, then tell me your result'; it added 199 'celebrity saju / compatibility with a celebrity' SEO pages for the long tail; and in 2025 it cloned the site for Taiwan (Traditional Chinese /tw) and Japan (/solo-jp). The checkout peaks three months ago and last month look like ad pushes.One-line growth story, compiled by us from public sources; speculation is marked separately below.
The main engine is paid ads: Facebook and Instagram feed ads bring in new customers, display ads are the third-largest channel, and nearly 20% of search is paid search.
The lubricant is free content: free Saju readings for yourself and celebrities and compatibility matching, so people start playing first and then pay for in-depth reports.
By estimate, about 37% of visits come from seeing ads; about 9 in every 100 visits click through to the checkout page.
Model: Lenny's racecar growth framework. Engine = paid ads (Facebook and Instagram feed ads and display ads bring in new customers, and report revenue buys more ads); lubricant = free Saju, celebrity compatibility pages and interactive Threads posts, getting people to play for free first and then pay for in-depth reports.
Real data: 677,600 visits, +11.35% month over month, direct as the largest channel at 34.18%, top three channel ranking (direct, organic search, display ads), organic search 80.37% and paid search 19.63% of search, 262 search terms, top search terms (the brand name "Tight Saju" in several Korean and Chinese spellings, and the generic term "Saju" in Korean), social sources (Instagram, YouTube, Facebook), countries (South Korea 63.74%, Taiwan 24.37%, US 3.62%, Hong Kong 3.41%, Singapore 1.77%), bounce rate 56.16%, 2.43 pages per visit and an average stay of 1 minute 49 seconds come from the public Similarweb page for last month; the shares of the second and third channels are not given on the public page. The 60,172 visits (about 60,200) to the Stripe checkout page come from quietmoney checkout page data, and the previous few months averaged about 10,600 per month. The older channel figures in the profile (search 49.4%, display ads 9.9%) use an earlier basis and are not used here.
Supporting evidence: Users on Taiwan's Dcard dating board say they saw someone sharing this site in Facebook ads; the site has the Facebook Pixel and GTM installed; in 2025 the official Threads account posted interactive posts "Enter your birthday and theirs, get your Saju compatibility instantly"; the sitemap has 199 Korean and 199 Traditional Chinese celebrity Saju pages; in 2025 it was copied to Taiwan (/tw) and Japan (/solo-jp). Local Korean payments go through PortOne/TossPayments and Stripe collects overseas payments, so checkout page visits do not represent all revenue.
Inference: The two channel circles are real, but how each splits is an inference: direct 231,600 = returning users 120,000 + saw a compatibility post or a friend sharing results 61,600 + opened it after seeing an ad 50,000; other channels 446,000 = saw an ad (display ads, paid search, searching the brand after watching an ad) 200,000 + searched Saju, celebrity compatibility 140,000 + saw a compatibility post 66,000 + returning users searching the brand name 40,000. The 2% click-through after seeing an ad, 10% after searching and 5% after seeing a post are guesses. Stripe payments mainly come from overseas users such as those in Taiwan, which is inferred from the profile.
Checkout: checkout.stripe.com
New-user source: SEO (Non-brand search). Search 49.4%, direct 34.2%, display ads 9.9%. Non-brand search is marked at about 49%, but the top terms are actually brand keywords (타이트사주, 緊緊八字) plus the generic term 사주. Social sources: Instagram, YouTube, Facebook. New users classified as SEO.
Top channel measured by Similarweb, others estimated by rank
| Attributable new users, last month | Share |
|---|---|
| SEO | 49% |
| Other | 51% |
Direct + brand search 34%. Direct and brand search count as viral only when the site shows sharing mechanics; the largest share wins, under 5% means Other.
| Country | Share of visits |
|---|---|
| Korea, Republic of | 63.74% |
| Taiwan | 24.37% |
| United States | 3.62% |
| Hong Kong | 3.41% |
| Singapore | 1.77% |
The homepage we captured is Traditional Chinese; most traffic is from Korea and search terms are mostly Korean, so there should be a Korean version (unverified).
Instagram, YouTube, Facebook
262 keywords · brand ~34% · organic 80.37% / paid 19.63%
| Similarweb top 5 keywords |
|---|
| 타이트사주 |
| 緊緊八字 |
| 타이트 사주 |
| 사주 |
| 사주타이트 |
Non-brand queries: 타이트사주, 緊緊八字, 타이트 사주, 사주, 사주타이트
| Category | Share of referrals |
|---|---|
| Computers Electronics and Technology - Other | 49.51% |
| Graphics Multimedia and Web Design | 5.38% |
| Tickets | 0.59% |
About 60,172 last month (Similarweb estimate), #32 of 606; the month before was 19,055. Opening checkout is not paying, so this is not revenue.
New-user source: SEO (Non-brand search). New users mostly come from need-based searches Search 49.4%, direct 34.2%, display ads 9.9%. Non-brand search is marked at about 49%, but the top terms are actually brand keywords (타이트사주, 緊緊八字) plus the generic term 사주. Social sources: Instagram, YouTube, Facebook. New users classified as SEO.
Not found (homepage shows 'no products at the moment').
Dossier confidence: medium. Parts 1 and 2 come from reading the site and Similarweb data; part 3 from public reporting, founder posts, Wayback snapshots, WHOIS and similar, with speculation marked.
The number: checkout visits is Similarweb's estimate of visits from the site to a checkout payment page; opening checkout is not paying, so this is not revenue. Small sites land on fixed buckets.
Data through last month. Page updated 2026-10-31.