130k installs, 90 ratings, 75 one-stars: Afterglow is a paid-acquisition business, not an AI video product

Afterglow: AI Video Maker (com.afterglow.motion) research report

Channel: Google Play | Developer: Matthew lee dorothy (gmail address m88128546@) | First released: 2026-09-16 | Category: Entertainment (ENTERTAINMENT). Data as of 2026-11-06. "Library" refers to our googleplay library (agent-im server), and "AppstoreSpy estimate" refers to historical data backfilled from app_day_ext.

Bottom line

This isn't an app that succeeded on its product. An anonymous small team or studio used a matrix of shell accounts plus paid acquisition to build an "AI video template plus gem (credit) consumable in-app purchase" shell product. Nearly all of its 132k installs came from paid traffic, and people who installed it mostly complain (75 of 90 ratings are one-star). Money comes back from "new users buying gems on their first purchase," not from retention. The "charts" it entered are the long tails of Entertainment grossing charts in Spain, Italy, Turkey and India, which have very low bars.

1. What the data itself shows

The install curve is "stepped" (AppstoreSpy estimates + library):

• 10-09 about 27k → 10-13 41k → 10-17 63k (+12,406 in a single day) → 10-21 79k (+13,675 in a single day) → 10-27 91k
• 10-28~10-29 about +2,10-30~11-03 per day, then back to 5,000–6,000,11-04 per day, ~05 drop to about 2,000,11-06, then 6,052 again.
• Stop-start, on-off: the classic shape of paid-ad budgets being switched on and off, not the smooth curve of organic word of mouth.

Charts: only appears in the Entertainment category charts. Its best rankings are Spain Entertainment Top Grossing 87–89, Italy Entertainment Top Grossing 98, Turkey 117–129, India 110–138, Mexico 96 (library chart_rank 10-29~11-06). On the day 11-06, Spain's Entertainment Top Grossing chart had only 118 apps in total, so the bar to reach #89 is low. It has never made any country's overall chart or the US chart.

2. User reviews: what users are complaining about (Play Store reviews, scraped with google-play-scraper: 11-06)

Almost all one-star, clustered around four issues:

• Gem paywall: "So expensive you can't even try it. You have to log in for ten days straight to earn enough gems to test the cheapest feature" (10-13); "Gave 10 credits and I can't do anything with them" (Italy, 11-02); "Not a single free prompt" (Spain, 11-02).
• Paid but nothing generated: "Bought gems, the video ran for over an hour and never stopped" (11-03); "Paid to make a video, it loaded for 3 hours" (11-03); "Topped up 24 hours ago, gems never arrived, support doesn't respond" (Spain, 10-18).
• Doesn't match the brief: "Ignores the template you pick" (11-05); "The output looks nothing like the description or the samples" (Spain, 10-09).
• "This app is about porn" (10-16). This suggests the ad creatives/templates lean suggestive, a common hook in AI video paid acquisition (specific creatives not verified).
• 5 five-star reviews consist of meaningless short phrases like "hi ga jd na ha ha ha bs GD GD GD", "polnito liko", "Very good". This suggests possible incentivized or fake reviews (not verified).

Bottom line: there is no retention to speak of. The revenue model is: ads bring people in → they see template showcase videos → they buy a first gem pack starting at $4.99 → disappointed, they leave.

3. Growth tactics: a sock-puppet account matrix + Meta paid acquisition

3.1 Sock-puppet account matrix (strong evidence)

The library turns up at least 3 packages sharing the same shell, with identical traits:

(library app / app_snap; creation date taken from the Published Time returned by the Jina crawl)

• Package names all follow <brand>.motion, all use developer accounts with "random English name + numbered gmail", all use a one-page "AI Cinematic / AI Motion Studio" landing page, all start at version 2.0.x (indicating the same codebase with a reskin), and the in-app purchase ranges are identical.
• Afterglow and Lumiora's install curves rise and fall in sync (AppstoreSpy estimates): 10-16 both nearly at 0 (+19 / +4), 10-17 both surge at once (+12,406 / +5,347), 10-20 both stop at once (+764 / +11), 10-21 both surge again at once (+13,675 / +11,547), 10-28~29 both stop again at once (+2 / +4). Same ad team, same budget switch.
• The purpose of the sock-puppet matrix: if one account gets banned by Google (with ratings and complaints like these, a ban is only a matter of time), the others are unaffected. It also spreads multiple brand names to capture the long tail of store searches for "AI video".

3.2 Meta (Facebook) ads

• There is a Facebook page named "Afterglow: AI Video Maker-1021-2" (page id 100095643424901): 0 followers, the About field lists a Vietnamese name "Điền Dung", the page still contains old images from 2021 years ago, and its only post was published on 2026-10-21 09:30. This is a typical "bought an old page, renamed it, and used it as the ad page" practice, and "1021-2" looks like a campaign batch number (the 21 batch, #2).
• And 10-21 was exactly the biggest single day of installs for Afterglow (+13,675) and Lumiora (+11,547).
• We could not scrape the specific creatives from the Meta Ad Library (it requires login to render). Our metaads / tiktokads / googleads watchlists don't track it, so the creatives, countries and ad count are unverified. But the page naming, the date match and the install shape together suggest Meta paid acquisition is the main channel, with the operator likely a Vietnamese team.
• Target countries can be inferred from the charts: Spain, Italy, Turkey, India, Mexico, Malaysia, France, the UK, Germany, Sweden, the Netherlands, Poland. These are all markets where CPI is much cheaper than in the US, and their Entertainment Top Grossing charts have a low bar, so a small number of first purchases can push it in.

3.3 Other channels

• TikTok (searched "Afterglow AI video maker" via TikHub): none of the first 20 results are content from this app. They're all AI short videos, songs, or "afterglow filter" tutorials sharing the same name. No creators or organic content.
• X (x-radar, last 60 days): 0 posts.
• Web: only auto-indexed APK sites like CNET Download, AppBrain, APKPure and applion. No press coverage, no Reddit discussion.
• Category choice of "Entertainment" over "Photo/Art & Design": the AI video leaders (PixVerse, Vidix, Nexus, etc.) crowd into Photo and Art & Design. The Entertainment category has less competition, making it easier to reach the category charts and gain exposure (inferred).

4. Competitors

Apps that turn "photos/prompts → AI video" are already a red ocean on Google Play (library app_profile 11-06):

5. Differentiation

Almost no product differentiation. The copy's "cinematic, controllable camera and motion" is packaging. User reviews say it ignores templates and takes hours to generate. Its only real "difference" from the similar shells is:

1. It sits in the Entertainment category plus small-language and European long-tail countries, where the Top Grossing bar is low;
2. It sells gems per pack (from $4.99) instead of a subscription, so the first-purchase barrier is lower than a $9.99/week subscription and it sidesteps subscription refunds and Google's subscription review;
3. Its brand names and landing pages look more polished than shells like FilmForge (Afterglow, Lumiora and StoryMuse names and pages are aesthetic), so ad click-through and store conversion may be higher (inferred).

These are differences in distribution and monetization, not in the product. It "broke out" (130k installs) on paid acquisition, not on differentiation.

6. Visibility and copycats

• Few people know about it yet: no public revenue, no media or big X accounts have mentioned it, zero TikTok/Reddit discussion, only pages on APK aggregator sites.
• Copycats: the business itself is mass replication within a big category. It is already one member of a matrix (Lumiora, StoryMuse), and the library holds more than a dozen same-style Entertainment-category AI video shells (FilmForge/Filmlet, HoneyDrip, CineDream, Framegen, Woolora, Aurion, MotoPic, etc.) from 8 to October. It's not that "someone is copying it"; everyone is copying the same model.

7. What others can borrow

• Category arbitrage: the same feature placed in a less competitive category (Entertainment). Category Top Grossing bars in small countries are extremely low (Spain's Entertainment Top Grossing has only 118 slots in total), so a handful of first purchases gets you onto the chart and into exposure.
• Per-pack credits instead of subscriptions: consumable in-app purchases starting at $4.99 have a low first-purchase conversion barrier, suiting a model where paid acquisition pays back quickly.
• Matrix + synchronized ads: the same codebase reskinned across multiple accounts and brands, so one ban doesn't sink the whole operation, and brand names and landing pages can be A/B tested.
• Cautionary tale: a 1.67 rating, almost all one-star reviews, and lots of "scam/refund" complaints. Accounts like this have short lifespans and high risk of Google bans or delisting. Whether it pays off depends entirely on whether first-purchase ARPU exceeds CPI. It isn't sustainable, and teams that want to build a lasting product shouldn't copy it.

8. Still unclear

• Real revenue: chart rank only shows a first purchase, the scale is unknown, and whether ROI is positive is unknown.
• Meta ad creatives, target countries and number of ads are unverified (the ad library didn't capture them; our ad library doesn't track this app).
• Whether the "porn" reviews point to the ad creatives or to in-app templates is unverified.
• Operator identity: the FB page points to Vietnam, but there is no further verified evidence; whether the portfolio has more packages not named .motion is not exhaustively checked.
• Which third-party video model the backend uses (presumably a third-party API; slow generation suggests a cheap queued channel) is unverified.

Sources

• Library googleplay: app, app_profile, app_snap, app_day_ext (AppstoreSpy estimate), chart_rank; query date 2026-11-06
• Play Store page and reviews: play.google.com/store/apps/details?id=com.afterglow.motion (scraped with google-play-scraper on 2026-11-06)
• Websites: afterglowmotion.com/, afterglowmotion.com/PrivacyPolicy.html; sister sites lumioramotion.com/, storymusai.com/
• Facebook page: www.facebook.com/p/Afterglow-AI-Video-Maker-1021-2-100095643…
• CNET Download:download.cnet.com/afterglow-ai-video-maker/3000-android-afte… ;AppBrain:www.appbrain.com/app/afterglow-ai-video-maker/com.afterglow.…
• TikTok search (TikHub) and X search (x-radar) 2026-11-06, no relevant content found