Global rank 90 in 30 days, but the website has nothing on it: haloreels.net is the backend of a short-drama app
What it is The official site and backend domain for Halo Reels, a free Android short-drama app (standard, Pro, and Max builds). The web page only shows contact info, terms, and a privacy policy.
Why it's climbing The rank is driven by the app: the three builds have 26.2M installs combined, constantly hitting subdomains like api, and that traffic gets counted toward the main domain. The app itself is pushed by a Facebook page and a set of sock-puppet drama packs bought in volume in the Philippines, Indonesia, and Latin America.
DHow you could play it There's nothing on the site to copy. What's worth learning is how to spot the signals (a homepage with only legal pages, almost no real visitors, ads that all jump to the store means look at the app) and the multi-build infrastructure: shared backend, one site template with the publisher name swapped.
Global rank 31k · 90 days ago, top 1M+
haloreels.net ranked outside the top 1M sites globally in August, then reached #30,580 by Nov 5. Open it and you only see a contact email, terms of use and a privacy policy. No player, no download button. The real growth is in the Android short-drama app Halo Reels: it ships as three packages with 26.2M installs in total, and every phone keeps hitting APIs under this domain in the background. This piece covers how to spot this kind of "fake website," what the playbook behind it is, and whether you should jump on it.
Nov 7: opening it, the homepage is one line of text
The title reads “HaloReels - Just watch it, it's HaloReels.” The body has three links, Contact Us, Terms, and Privacy, plus a feedback email, [email protected], and a footer credit “2026 HALOREELS LLC.” There's only a single banner image and not a single link to Google Play. No robots.txt, no sitemap, no search optimization at all.
Three things in the page source are worth noting:
- The keywords field stuffs in competitors' names: “reel short,” “drama box,” “short max,” “dreame short,” “reelbox,” “alta tv,” and more than thirty others, all other short-drama app names.
- The first font in the stylesheet is “PingFang SC,” Apple's Simplified Chinese font. The site was most likely built by a Chinese-speaking team (inferred).
- The same site also hosts index-pro.html and index-max.html, with identical templates and only the footer credit changed to “HALOREELS PRO LLC” and “HALOREELS MAX LLC.” Google Play requires each app to have a developer website and privacy policy, and these three pages are what gets submitted for store review.
The domain's certificate was issued in Jul 21, and the basic app was live by Jul 24. So this is a fresh domain bought for the app, not an old domain that changed hands.
Why the rank rose: phones hit it in the background
The change in global website rank (Tranco) follows the same curve as the install numbers for the three builds:
| Date | Global rank | The app on Google Play at the same time |
|---|---|---|
| 08-20 | 272,965, first entered the top 1M | Base version 08-16 hits 1M installs |
| 08-27 | 124,425 | Pro and Max plans 08-18 launched |
| 09-10 | 53,945 | Pro passed 1M on day one |
| 09-24 | 45,831 | Max 09-20 hits 1M |
| 11-05 | 30,580 | Three builds total 26.2M installs, +3.03M in one week |
Part of this rank comes from DNS lookups, and subdomains like api.haloreels.net get counted toward the main domain. Every time the app loads episodes or reports data, it adds to haloreels.net's rank.
Two more things show it isn't really a website:
- Real people rarely open it in Chrome. In Chrome's real-user stats (CrUX), it never makes the global top 1M. It only reached the top 500k in the Philippines in 8 and the top 1M in Indonesia in August. Those two countries are the app's main markets, so the visitors are most likely people who tapped the privacy page inside the app, or who searched for the app name and landed on the site (inference).
- Ads don't send traffic to the site. The Facebook page running its ads is called "PassionPro-Short". All 112 ads with visible links go straight to Google Play. Not one of them lands on this site.
So the site's ranking is just a shadow of the app's growth. The website itself has no users.
Behind it sits a row of sibling domains
Check where api.haloreels.net points (11-07): three AWS Singapore servers. The api.reelshortfly.com of another short-drama app, ShortFly, points to the same three servers. The data-reporting subdomain teapi for both apps points to the same Alibaba Cloud Singapore load balancer. There is also one IPv6 address hosting haloreels.net, reelshortfly.com, anyshorts.net, dramafuture.com, and reeljoys.com together.
These sibling domains also climb the global rankings together (11-05): passion-drama.com 36,377, reelshortfly.com 45,791, dramlyol.com 67,612, bonusdrama.com 89,821, anyshorts.net 136,049. Each one backs a short-drama app registered under a different shell developer, yet all their ads are run by the same Facebook page with only 4,700 likes. In Nov 6 it was running 442 ads, and 20,301 in total historically, in Spanish, Indonesian, Portuguese and English.
To switch to a new front, all you need to do is buy a domain, change the package name, and drop a website template on it, while the backend connects to the same servers. That's how one operator can roll out ten packages at once. A Chinese going-global short-drama daily attributes this group to a single Shanghai team. The third-party chart Insightrackr reports that in September, Halo Reels Max downloads rose 2598% month over month, almost all from Meta's Audience Network (ads shown inside other apps).
Where the money comes from
The money isn't on the website, it's in the app: an ad plays before each episode, and watching an ad unlocks the next one, or you buy a weekly or annual pass, listed in-app at $7.99–29.99. The main markets are the Philippines, Indonesia, Thailand, Mexico and Brazil, where installs are cheap and users are used to "watch ads for free". Revenue isn't public. How it converts inside the app and what reviewers complain about, we cover in the Halo Reels Pro piece.
Who it's competing with
The big players in short dramas, ReelShort and DramaBox, both have playable web versions, each with over 300M installs (Google Play, 11-06). On the free side there's Kunlun's FreeReels (246M), ByteDance's Melolo (147M), and TikTok's PineDrama (84M). Halo doesn't even have a web version, which shows it never planned to get users through the web. Its rivals are the same wave of "shell developer + template website" drama packages. More than a dozen of these exist, including its own siblings.
If it were you
One-line verdict: D. There's nothing to copy at the website layer. Its ranking is driven by the app backend. The barrier to this playbook of "one homepage pushing a line of shell packages" is budget, batches of developer accounts and ad-buying experience. With more than a dozen shells already running it, it isn't worth it for solo indie devs to follow.
Who it's for: people who watch website rankings and look for "what's rising". This piece helps you avoid misreading one kind of signal. Teams already mass-producing going-global apps, who need a website and privacy page for each package, can also borrow its infrastructure.
Who it's not for: anyone who sees "ranking up 30x in 90 days" and wants to build a similar website. There are no web users to chase here.
How to do it: two things. First, recognise the signal: if a domain's ranking surges, its homepage has only terms and privacy, it's absent from CrUX global data and only shows a little in one or two small countries, and all its ads jump to the app store, then it's almost certainly the backend of some app. Go look at the app instead. Second, if you want to build a shell-package matrix, copy the foundation: one API and data-reporting backend, with subdomains split by product; one static website template, where swapping the footer gives you a new package's "developer site"; and spend all acquisition money on Meta ads that jump straight to the store.
How to acquire users: the website itself acquires nothing. Acquisition runs entirely through Meta feeds and Audience Network, using high-drama clips, with language swapped by country, linking straight to Google Play.
Timing: not applicable. The website has no window. The app playbook's window is measured in months: once Google Play starts cleaning out duplicate short-drama packages in bulk, this group of packages and their domain rankings will fall together.
What would show I got it wrong: if haloreels.net one day launches a playable web version and starts doing search optimisation, or CrUX shows it entering the global top 1M, that means it has begun taking the web seriously, and it should be reassessed as a web product.
Research notes: competitor table, acquisition evidence, full research report
- Competitors
Name Where Launched Scale How it differs from this one ReelShort iOS 2022-08 Play 307M installs, 7 days +5.51M Paid unlocks and original series, with playable web versions. Halo is free and Android-only, its website just a shell. DramaBox iOS 2023-04 Play 307M, Top Grossing #6 Mainly paid, heavy on licensed content, both platforms plus web FreeReels / DramaWave (Kunlun Tech) Android 2024 246M / 152M installs The big-company benchmark for free short dramas with a large library. Halo grows through multiple packages bought with paid acquisition Melolo (ByteDance) Android 2024-11 147M installs Also free, focused on Southeast Asia, fewer ads PineDrama(TikTok) Big-company features 2025-12 84M installs Has TikTok's own traffic. Halo relies entirely on paid acquisition Free short dramas from shell accounts like HotMini Android 2025-05 55.9M installs Predecessor using the same playbook, but not split into multiple packages - Where users come from
- Verified Meta ads (mainly Audience Network), jumping straight to Play, bypassing the website: metaads library PassionPro-Short (4,700 likes, 442 ads running, 20,301 total historically). All 112 ads with links landed on play.google.com, 27 of them pointing to the three Halo packages. Insightrackr's September chart states the traffic came via Meta Audience Network.
- Verified Multi-package, multi-shell matrix sharing one backend: api.haloreels.net and api.reelshortfly.com share IP 52.74.107.8. The teapi subdomains of both CNAME to the same Alibaba Cloud Singapore ALB. The same IPv6 hosts 5 sibling domains (bgp.he.net).
- Verified A Facebook page posts episode clips to drive traffic: a Facebook page named Halo Reels posts videos with the caption "Full version on comments".
- Inferred SEO / organic web traffic: the site has no robots file, no sitemap, and no content pages. It is absent from CrUX global top 1M.
- Inferred TikTok / X organic content: x-radar shows 0 posts. TikHub returns no official accounts (see the companion Play report).
Why haloreels.net (Halo Reels) shows up in the rankings
Channel: web (Tranco rise recall, pool C1) | Research day: 2026-11-07 | Companion reports: three in the same directory, play_com.halo.reels.movie.fast / .pro / .max (all app-side details are there)
Bottom line
haloreels.net is not a web product. It is the website and backend domain for the free Android short-drama app Halo Reels (three packages: basic, Pro and Max, under the shell developer Nutri Plaza). The homepage has only a footer, contact info, terms and a privacy policy, with no player and no download button. All Meta ads land directly on Google Play, bypassing this site. Its Tranco rank climbed from outside the top million to 30,580. The source is the three packages, which together have 26.2M installs. The installed apps keep requesting api.haloreels.net and teapi.haloreels.net (Tranco merges subdomains into the registrable domain, and DNS-based data sources count all of those requests). So nearly all the traffic comes from inside the app, and the web side can be ignored. The app itself grows through a playbook run by a Chinese going-global short-drama team (recorded by industry dailies as the "Shanghai Jinrong Youxi group"): a shell-package matrix + Meta Audience Network paid acquisition + "all episodes free" in emerging markets. The product itself has no differentiation.
1. What the website itself is
- The homepage www.haloreels.net is titled "HaloReels - Just watch it, it's HaloReels". The body only has Contact Us / Terms / Privacy, feedback email [email protected], and the signature "2026 HALOREELS LLC". The HTML contains only a banner image and a few internal links, with no Play link and no web playback (2026-11-07 fetched it with Jina and curl).
index-pro.htmlandindex-max.htmlare the same template, signed "HALOREELS PRO LLC" and "HALOREELS MAX LLC" respectively, serving as a "developer website + privacy policy" for each of the three packages on Play.- The page last-modified date is 2026-08-13. The certificate 2026-07-21 covers
haloreels.netand*.haloreels.net, and 07-23 coversteapi.haloreels.net(certspotter). The basic app was listed on 07-24, which fits: the domain was registered new for the app, not an old domain changing hands. - There is no robots.txt and no sitemap. No SEO work has been done at all.
2. Where the traffic comes from: the app backend, not the web
Tranco rank (traffic library, rank_day):
| Date | Rank | The app's status over the same period (googleplay library, see companion report) |
|---|---|---|
| 08-20 | 272,965 (first entered the top 1M) | Base version 08-16 hits 1M installs |
| 08-27 | 124,425 | Pro / Max listed on 08-18 |
| 09-10 | 53,945 | Pro passed 09-10 1M |
| 09-24 | 45,831 | Max began running ads on 09-12; 09-20 passed 1M |
| 10-15 | 37,341 | |
| 11-05 | 30,580 | Three packages combined reached 26.2M installs, 7 days +3.03M |
The rank curve and the install curves of the three packages move in sync. Tranco is a 30-day rolling average and lags by 1–2 weeks, which fits.
Evidence: the requests hit API subdomains
api.haloreels.netresolves to AWS Singapore IPs (52.221.27.59 / 52.74.107.8 / 52.221.161.86), and the root path returns the default nginx page.teapi.haloreels.netCNAMEs to an Alibaba Cloud International Singapore ALB (alb-1y2ts4nzavggt4d7io.ap-southeast-1.alb.aliyuncsslbintl.com), and returns "Cannot GET /" (a Node backend).teapiis probably the data-reporting endpoint of ThinkingData (a Shanghai company whose SDK is called TE), but this is unverified.- Tranco folds the subdomains into haloreels.net, and DNS data sources in the mix, such as Cisco Umbrella and Cloudflare Radar, count the app's backend requests as "popularity". The recall card also flags this: "Not in CrUX global top 1M, the rise may come from DNS / API traffic".
- CrUX shows real Chrome visits are very few: only
https://www.haloreels.netentered the top 500k in the Philippines in 2026-08 and 09, and the top 1M in Indonesia in August. It is absent from the global top 1M (traffic library, crux_domain). This is probably users who opened the privacy / terms / contact pages from inside the app, or who searched the app name and found the site. The volume is small and matches the app's main markets (the Philippines and Indonesia). - Meta ads don't land on the official site: the PassionPro-Short homepage in the metaads library (page_id 1156915574179436) runs ads that all send traffic to play.google.com (112 linked ads in total, 27 of them pointing to the three Halo packages). landing_domains is only play.google.com. The googleads library shows no_ads.
Sister sites are climbing too, a sign this is a shared pattern across the portfolio (traffic library, 11-05 rank): passion-drama.com 36,377, reelshortfly.com 45,791, dramlyol.com 67,612, bonusdrama.com 89,821, anyshorts.net 136,049, nicetvs.net 140,226, dramafuture.com 146,520. These are all the same template of short-drama front sites (see section 2 of the companion report). The infrastructure is shared too:
api.reelshortfly.comandapi.haloreels.netshare IP 52.74.107.8;teapi.reelshortfly.comandteapi.haloreels.netCNAME to the same Alibaba Cloud ALB;teapi.passion-drama.comandteapi.dramlyol.comshare another ALB (from this round of dig).- Hurricane Electric's reverse-IP page shows that the same IPv6 address hosts anyshorts.net, dramafuture.com, haloreels.net, reeljoys.com and reelshortfly.com (bgp.he.net).
This is the strongest technical evidence so far that these sites share one operator, stronger than the footer credits or the industry daily's attribution.
3. Why users come, stay, and pay (app side, from the companion report)
- Where users come from: paid Meta acquisition, mainly Audience Network. PassionPro-Short has only 4,700 likes, 11-06 is running 442 ads now and 20,301 in total, and it promotes 10 shell apps, with copy switched to Spanish, Indonesian, Portuguese and English (metaads library). Insightrackr's September chart says Halo Reels Max downloads rose 2598% month over month and it entered the overseas short-drama download ranking at #7, "centralized user acquisition through Meta Audience Network, with AI-generated human-like drama to boost emotional engagement" (Zhihu repost). A Facebook page called Halo Reels also posts episode clips captioned "Full version on comments" (FB video), which is also used to drive traffic.
- Why now: in 2026 overseas short dramas shifted from pay-to-unlock to free plus ads, and three of the seven titles that passed 100M downloads are mainly free (Wall Street CN). Ad costs in Brazil, Mexico, Indonesia, the Philippines and Thailand are cheap, and rewarded ads pay back quickly.
- Why users stay: the most consistent praise in user reviews is "full episodes free, nothing locked."
- Why users pay: each episode has 1–3 ads inserted, pushing users to buy the weekly VIP; in-app purchases run $7.99–29.99. The entertainment top-grossing chart ranks it 55–95 in Southeast Asia and Latin America (Google Play chart_rank).
4. Competitors
| Competitors | Where | Launched | Scale | How it differs from Halo |
|---|---|---|---|---|
| ReelShort | iOS + Android + web | 2022-08 | Play 307M installs, 7 days +5.51M | Pay-to-unlock, original productions, brand disclosed, with a playable web site. Halo is free, Android only, and its official site is just a shell. |
| DramaBox | iOS + Android + web | 2023-04 | Play 307M, Top Grossing #6 | Mostly paid, heavy on licensed content. |
| FreeReels / DramaWave (Kunlun Tech) | iOS + Android | 2024 | 246M / 152M | The big-company benchmark for free short dramas with a large library. Halo grows through multiple packages bought with paid acquisition |
| Melolo (ByteDance) | Mainly Android | 2024-11 | 147M | Also free, focused on Southeast Asia, with fewer ads; reviews often compare it to Halo. |
| PineDrama(TikTok) | Android | 2025-12 | 84M | Has its own TikTok traffic. |
| Free short dramas from shell accounts like HotMini | Android | From ¥2025-05 | 55.9M | A predecessor using the same playbook. |
| halodrama.cc(Halo Drama) | Web / App | Not verified | Not verified | Another short-drama site with a very similar name; relationship unverified. |
(Scale from the googleplay library 11-06, see the companion report.) The big players in the "short-drama web site" space all have playable web versions. Halo has no web version at all, which itself shows its playbook is pure app-based paid acquisition.
5. Differentiation
- Website: no differentiation, and it doesn't carry acquisition.
- Product: the same free vertical short-drama format as Melolo and FreeReels, with heavier ads. The "AI human-like drama" mentioned by Insightrackr may lower content costs (unverified; the ai_media flag on this page in the metaads library is 0).
- What actually sets it apart is operations: one brand split into three packages, multiple shell developer accounts, one Meta page buying ads for all of them, and backend, site templates and servers shared across a dozen or so shells (the DNS evidence above). Cloning a new shell only takes a new domain, a new package name and the same template. That's what made it work, not the product.
6. Visibility and copycats
- Chatter in small circles: a Chinese short-drama daily for global expansion (reposted by amz123 from shortdramadata.com, which attributes it to "Shanghai Jinrongyou / Nutri Plaza shell") and Insightrackr's monthly chart have named it; English media, X (x-radar 0 posts) and TrustMRR have nothing, and revenue is not public. What you can find online is only APK sites, scamadviser and Cloudflare Radar domain pages.
- Copycats: the operator is copying itself (BonusTV, ShortFly, Drama World, NiceTV and Passion Drama all use the same template and the same backend); on Play there's a name-squatting package com.haloreelspro.appmoviesreels; and the industry has plenty of free short-drama shells.
7. Lessons for our recall rules
This case is a textbook example of an app backend pushing a website domain into Tranco. Signals to watch: no CrUX data globally, only low-tier data in the main app markets; no functionality on the homepage; ad landings all go to the store; sibling domains on the same template rising together. When you see "Play developer site + CrUX only in small countries at low tier" for a C1, go check the app directly instead of researching it as a web product.
8. What to borrow
- How a multi-package portfolio builds its infrastructure: share one API and analytics backend (split into subdomains per product), share one static site template (index / index-pro / index-max, three pages matching three LLC footers), so the marginal cost of launching a new package is almost zero.
- The official site only provides what store review requires (developer site, privacy policy, contact email). It does no SEO and no web playback; all acquisition money goes to Meta ads that jump straight to the store.
- The rest (Audience Network acquisition, free full episodes in emerging markets, rewarded ads plus weekly subscriptions, multiple packages taking chart spots) is covered in the companion Play report.
9. Uncertain
- Whether
teapiis ThinkingData's reporting endpoint: inferred only from the name and response characteristics, not verified. - Where the small amount of real-person Chrome traffic in CrUX comes from (privacy pages opened inside the app, search, or something else) is unknown; there is no path data.
- The equity relationship between "Shanghai Jinrongyou" and HALOREELS LLC / REVERIE STUDIO LLC / 劇飛有限公司 is unverified; registration data (RDAP) was not found.
- Ad spend, ROI and revenue scale are unknown; whether the content is AI-generated, licensed or repurposed is unclear.
Sources
- traffic library (run by the agent-im machine, 2026-11-07 lookup): rank_day (haloreels.net and 7 sibling domains), crux_domain, product_domain
- metaads library: the advertiser for Page 1156915574179436 "PassionPro-Short", and its ads (landing domains, links)
- Site captures: www.haloreels.net/, /index-pro.html, /index-max.html (Jina Reader + curl, 2026-11-07)
- DNS: dig on the api / teapi subdomains (haloreels.net, reelshortfly.com, passion-drama.com, dramlyol.com); certificates: certspotter API
- ipv4.bgp.he.net/ip/2406:da18:4c2:e701:9145:e5d8:2c8e:44ca (sibling domains on the same IP)
- Insightrackr September chart: www.zhihu.com/pin/2092923930403135633; track: wallstreetcn.com/articles/3783400
- Facebook promo video: www.facebook.com/61592931008611/videos/full-episodes-unlocke…
- App-side data and sources such as the amz123 daily: same directory, play_com.halo.reels.movie.fast.md