Chinese-language backlink/traffic data rankings for AI tool sites (Columbus AI); users look up AI products' data rankings and backlink profiles.
A Chinese indie developer with 8 years of scraping/reverse-engineering experience (X handle @jinlv11, "Columbus") scraped the intel everyone in the "AI tool sites for overseas markets" crowd wants (which AI sites are growing, which are collecting payments, what backlinks and keywords they use) and packaged it as a paid ranking. On launch day around 2026-08-03, Chinese go-global KOLs with 10k–70k followers (@hezhiyan7, @GoSailGlobal, @guishou_56, @meepodev, @AI_Jasonyu) were already posting "wealth code" tweets with?aff= 30%-off referral codes, reaching 2k+ UV in 3 days. Since then it has retained and expanded this base of Chinese go-global developers through near-daily tweets about "ranking updates / new features / new models", a "Twitter daily digest" WeChat group for subscribers, and MCP/API access plus Alipay payments.One-line growth story, compiled by us from public sources; speculation is marked separately below.
The main engine is overseas-focused KOLs reposting with referral codes: from launch day at least 5 Chinese-language going-global creators with 10,000 to 70,000 followers posted "wealth code" tweets on X with a 30% off referral code. Direct visits are 62%, and organic social, sourced from X, is second.
The lubricant is the author posting updates almost every day: rankings, new features and new models in rotation, a subscription that includes a "Twitter AI daily" WeChat group, plus Alipay and a one-time pass.
By estimate, over 40% of visits are subscribers coming back to check the rankings; the visit count is an estimate, and about 33 of every 100 visits click through to the checkout page.
Model: Lenny's racing growth framework. Engine = overseas-focused KOLs reposting with referral codes ("wealth code" tweets plus a ?aff= 30% off code, with the product itself serving as KOL content material); lubricant = the author posting updates almost every day, a subscription that includes a "Twitter AI daily" WeChat group, and Alipay and a one-time pass that lower the barrier to pay.
Real data: +13.04% month over month, largest channel direct at 62.31%, top three channels ranked (direct, organic social, referrals), organic search 100% of search, 16 search keywords in total, top keywords (columbus.tools, "Columbus ai", "Columbus ai backlink ranking", "Columbus ai product", "Columbus.tools ai"; the last four are Chinese-language keywords), social source X, countries (China 95.5%, Japan 4.5%), bounce rate 66.9%, 3.04 pages per visit, average time on site 2 minutes 37 seconds, global rank #3263946, all from the Similarweb public page for last month; the public page gives no total visits, does not say how much the second and third channels have, and has no referring sites. The 8,596 visits to the Stripe checkout page (checkout.stripe.com) are from quietmoney checkout data. The "social 28.3%, referrals 5.7%" in the profile is an earlier basis and is not used here.
Supporting evidence: The author @jinlv11 claims 8 years of web scraping and reverse engineering experience, and the domain was registered in 2026; the official account said UV was 2k+ three days after launch; @hezhiyan7 (16,600 followers) 12,900 views, @GoSailGlobal (39,000 followers) 20,000 views, @guishou_56 (11,000 followers) 26,000 views, @AI_Jasonyu (72,000 followers) 20,000 views, with 3 of them carrying a ?aff= referral code; a @meepodev tweet says it makes "$10K a month" (a third-party claim, not verified); it later added a revenue ranking, MCP access, Alipay payment and a one-time pass, and a pixel-perfect clone site also appeared.
Inference: The total visits estimated from global rank #3263946 is 8,300 (fitted from 437 sites with real traffic, where half the sites are within 17% error), which is even lower than the 8,600 checkout page visits, so total visits use checkout page visits x 3 = 25,800 instead, with an error that could exceed double; both channel circles are marked as approximate (~), and the channel shares and ranking are real. How each source splits is an inference: direct 16,100 = subscribers coming back 10,000 + group members and peers recommending 2,600 + KOL tweets 2,500 + the author's tweets 1,000; other channels 9,700 = clicking links in KOL tweets 4,500 + clicking the author's tweets 2,500 + subscribers searching the brand name 1,500 + clicking links after being recommended 1,200. By rank organic social is second, so the two X paths get more. Checkout page visits of 8,600 are more than the 8,500 that keep browsing, so only one step is drawn after arriving. That 3% of those who see a tweet click through and 30% of those recommended it visit are guesses. Payments through Alipay do not go through Stripe and are not counted in checkout page visits; the 8,596 recorded by quietmoney is exactly the same as for another 130 or so sites in the input pack, suspected to be a tiered estimate from the data source.
Checkout: checkout.stripe.com
New-user source: Social media (Organic social). Direct 62.3%, social 28.3% (X-twitter), referral 5.7%; search terms are essentially all brand keywords ("哥伦布 ai", "哥伦布 ai外链榜单"). New users come mainly from the founder's content on X; existing users return directly.
Top channel measured by Similarweb, others estimated by rank
| Attributable new users, last month | Share |
|---|---|
| 28% | |
| Other | 72% |
Direct + brand search 62%. 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 |
|---|---|
| China | 95.5% |
| Japan | 4.5% |
Chinese (inferred; the site is blocked by Cloudflare and couldn't be verified)
X-twitter
16 keywords · brand ~62% · organic 100% / paid 0%
| Similarweb top 5 keywords |
|---|
| columbus.tools |
| 哥伦布 ai |
| 哥伦布 ai外链榜单 |
| 哥伦布 ai 产品 |
| 哥伦布.tools ai |
Non-brand queries: 哥伦布 ai, 哥伦布 ai外链榜单, 哥伦布 ai 产品, 哥伦布.tools ai
| Category | Share of referrals |
|---|---|
| Social Media Networks | 3.68% |
| Computers Electronics and Technology - Other | 0.61% |
About 8,596 last month (Similarweb estimate), #382 of 606; the month before was 0. Opening checkout is not paying, so this is not revenue.
New-user source: Social media (Organic social). New users mostly come from YouTube, X, Reddit, TikTok and the like Direct 62.3%, social 28.3% (X-twitter), referral 5.7%; search terms are essentially all brand keywords ("哥伦布 ai", "哥伦布 ai外链榜单"). New users come mainly from the founder's content on X; existing users return directly.
Pro $29/month; $199/year
Dossier confidence: low. 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.