No need to join dozens of Facebook groups anymore: a small team has taken Taiwanese fan-merch trading and folded it into an app with "catalog + escrow"

KOCA (Kacard Universe) is a secondhand marketplace app for K-pop photocards and ACG merch, built by a small team in Taipei. It launched in April 2025, and by November 2026 it had 7,479 ratings on the Taiwan App Store with an average of 4.83, over 81k installs on Google Play, and the company says it has more than 240k registered users. It spent little on paid acquisition. What it did was simple: it moved photocard trades, which used to be scattered across dozens of Facebook groups where cards were hard to find and scams were common, into an app where every card has an entry and the platform holds the money. Below, we explain how it took off, where the money comes from, who it's competing with, and whether you could build something similar elsewhere, in a different category.

An FAQ seen by 64k people

In April 2025 23, the day after KOCA launched, the official account posted a long Q&A on Threads. The first question was a user's own words: "I'm just selling a photocard, why do I need to give so much personal information?" The answer was blunt: the team had debated whether to get involved in trades at all. Just exchanging information, like the Facebook groups, "is definitely the simplest and least trouble for the team," but "we don't think that actually solves the problem of photocard trading." Handling money means following Taiwan's anti-money-laundering rules, so sellers have to verify their real identity.

The second question was harsher: "I've never even heard of your company, why should I hand my data over to you?" The official reply gave the company's unified business number and said the team includes a former technical lead from inline, a designer, a senior back-end engineer, and "fan operations staff." The third question was an admission of fault: early identity verification was done by manual email review, and "we didn't expect so many users to pour in early on, and that's really our mistake." The closing lines were "thanks to the hundreds of users who trusted us early on," and "the Android version will launch within a month."

As of today this thread has 64k views (on Threads, 11-06). A year and a half later, the "hundreds" from back then has become an official count of 240k registrations.

What it is

A photocard is a random member photo included with a K-pop album, one per album. To collect a whole member you have to trade for or buy them on the secondhand market. In Taiwan, that market used to run almost entirely through Facebook trading groups and LINE chats: each group had one or several Facebook groups to join, dozens in total. Searching by card name didn't work, payment was by bank transfer, and posts about scams, where the other person took the money and didn't ship ("runners"), were common on Dcard's fan boards. General e-commerce sites had some sellers too, but as one user review put it: "Before, buying photocards meant going to X (Twitter). Prices were high and there was no way to tell if they were real."

KOCA did two things:

• The catalog. Each card gets an entry with historical sale prices and the current lowest asking price. You can search by image, list in one tap, and get notified when a card arrives or drops in price. Sellers who don't know how to price can open a bidding auction. The catalog wasn't entered by the team. It was built by volunteer users called "Star Planet Captains." It reached 4k entries in the first 70 months and over 1.26M entries by November (the official Instagram cites about 2025-08; NOWnews, 11-03).
• Escrow. Sellers verify their real identity with an ID card and a face scan. Buyers' payments are held by the platform, and the seller is paid only 24 hours after the buyer receives the item. Payment is cash-on-delivery at convenience stores, shipping runs through 7-11, and each order gets a virtual account, so sellers never have to disclose their bank account.

The first line on the homepage is "No need to join dozens of groups anymore."

How fast it grew

• iOS (Taiwan): listed on 2025-04-15, 7,479 ratings, average 4.83. In the last 7 days it gained 209 ratings, roughly 30 per day. It ranked #24–#39 on the Taiwan shopping free chart (App Store, 10-28 to 11-06).
• Android: 80,869 total installs worldwide, +1,873 over the last 7 days, running at a few hundred a day. It ranked #24–#34 on the Taiwan shopping free chart and only charted in Taiwan (Google Play, 10-29 to 11-06). The store page lists the launch date as 2026-08-25, but the earliest review is from May 2025, the date of the relisting.
• Official figures: less than 4 months after launch, GMV passed NT$1.5M and 10k items sold (official Instagram). Press reports from late October to early November 2026 cite monthly active users of 160k to 200k and 240k registrations (Storm Media 10-24, NOWnews 11-03, with the two figures differing).
• Team: LinkedIn lists only one co-founder, and the company calls itself "a small team beyond anyone's imagination." The app is now on version 7.76, with recent weekly updates.

Comparing review counts on the Taiwan App Store: KOCA has 7,479, the game-card app Kapapa has 1,165, and Korea's Pocamarket has only 4. Within Taiwan's fan photocard niche, it's essentially the only one.

Android's daily installs over the past week fell from over 700 to around 200. Store install counts update only every few days, so part of this gap is a statistical artifact. Whether growth is slowing needs a longer data window (inference).

Where users come from

The official Threads account. Taiwan's young people and fan communities are mainly on Threads, and KOCA's official account @koca.universe has 28.6k followers. Its content falls into three types: a monthly "fastest-selling photocards" ranking (the October post got 3.1k likes and 783 reposts), public action against bad sellers with permanent bans (6.3k likes and 1.7k reposts), and long-form direct responses to criticism like the one at the top. Its persona is a customer service rep who talks back, and that has spread widely in fan circles.

Word of mouth. Someone on Threads wrote, "This should be everywhere. The photocard encyclopedia is amazing but hardly anyone knows about it." In the groups, people started replying, "Photocards can be sold on KOCA, it has platform protection." The official site put five-star store reviews straight onto its homepage.

Let users build the encyclopedia. Starting in May 2025, KOCA recruited "Planet Captains": each idol's "planet" needs at least 3 people who can add and maintain card entries. On FB they publicly praised "the men who have built the most encyclopedias." To complete their own idol's encyclopedia, Captains pull in fellow fans, and the encyclopedia pages get indexed by search engines, so many card pages under koca.shop show up in search.

Offline pop-up. They ran a pop-up in the East District underground mall in Taipei, invited sellers to set up stalls on site, and the live card draw for girl group cards ran out within an hour. The official T-shirt doubled as a "permanent offline VIP pass."

Opening one "planet" at a time. It expanded from K-pop to Ijiwa Kawa (Chiikawa), Sanrio, Haikyu!!, esports teams (T1, HLE, GEN.G), game card games, CPBL baseball, F1, and more than twenty planets in total, and split out two sub-brand accounts, @koca.acg and @koca.tcg. Each new planet brings a batch of people who used to trade in group chats. Geographically it has opened to Hong Kong, Macau, Singapore and Malaysia, but Hong Kong's iOS app has only 22 ratings, so it hasn't taken off there the way it did in Taiwan.

Not paid. I found no evidence of paid acquisition, and searching TikTok turns up no related videos. The two Fengchuan Media and NOWnews articles about "netizens pushing it hard" are both captioned "Image provided by KOCA," which means they were brand placements after growth started, not the cause of it.

Where the money comes from

For the first few months after launch it charged zero fees, which was how it pulled sellers over from the groups. Now it charges sellers a 2% transaction fee, plus 3% on card payments, and rules forbid sellers from passing fees on to buyers (user reviews, store page). Based on the official figure of "GMV of NT$1.5M in the first 4 months," that was only about NT$400K in monthly volume at the time. Current GMV and fee revenue are not public.

Store reviews from the last two months show it shifting from "grow big first" to "collect money and enforce order," and sellers are unhappy: "The fees are about to kill us," "You charge money to unlock a banned account, just go rob people," "The official side doesn't care about sellers at all... they just rule that the buyer is right," plus several reports of crashes and forced logouts after updates (App Store Taiwan reviews, September to November). Some defend it: "I see lots of people online (especially sellers) criticizing how KOCA handles things, but if you read the content carefully the official side is completely right" and "They enforce strictly and don't worry about piracy, so it feels safe." The average rating is still 4.83.

Who it's competing with

In 2026 a few photocard apps newly appeared on the Taiwan App Store, all with ratings near 0, and not one has yet grown into a followed category. Locally in Taiwan, KOCA already has 1.26M entries in its encyclopedia and the largest seller base, so it's very hard for a latecomer to challenge it head-on.

If it were you

Verdict: Worth trying, but don't go after KOCA in Taiwan. What KOCA proves is a formula you can take elsewhere: a collecting scene where trading still lives in FB/LINE groups, product is non-standardized, and scams are common, then build "community-built catalog + escrowed payments + local shipping," which can grow without paid acquisition. The opportunity lies in other regions and categories. The weakness is that this is a two-sided trading platform. Cold start, payment compliance and dispute support are all unavoidable, which makes it much heavier than building a utility app.

Who it's for: People already inside a collecting scene who know the group admins and the big sellers; small teams that can handle local payment and shipping integrations (convenience store pickup, cash on delivery, e-wallets); people with e-commerce or community operations experience.

Who it's not for: Indie developers who only want to write code and don't want to handle disputes like "the seller says the card is fine, the buyer says it's creased." Disputes and customer service are the bulk of KOCA's current bad reviews. Also not for people hoping to monetize quickly with a tool; platforms like this need zero fees for the first few months, so there's almost no revenue until volume picks up.

How to build it: Pick a place where Pocamarket and KOCA haven't seriously built local C2C. K-pop second-hand photocards in the Philippines and Indonesia still trade in Facebook groups via GCash and Shopee cash-on-delivery. The only tools are collection trackers like Bibliocards, with no local marketplace that offers escrow (web search, 11-06). Or change category: player cards, blind boxes, merch, esports gear. The test is whether the three conditions hold: it's hard to search for, people fear getting scammed, and trading runs through groups. For v1, cover one group or one IP as a catalog, and have the most active fans act as moderators to fill in the listings. Hold the money in escrow and release it only after delivery is confirmed. That step is the core of the product, so don't cut corners on it.

How to acquire users: Open an official account on the platform where your target users spend the most time (Threads in Taiwan; in the Philippines and Indonesia, most likely FB and TikTok; this is an inference), and copy KOCA's three moves: a monthly trading leaderboard, public handling of scammers, and direct responses to criticism. Find admins and big sellers in the groups, and in the first few months waive fees to get them to move over. Run one or two offline swap meets to turn familiar faces into your first seed users.

When to do it: KOCA has already expanded to Singapore and Malaysia, and Pocamarket is spreading into Southeast Asia and Hong Kong. Once one platform in a region gets its encyclopedia to hundreds of thousands of entries, latecomers find it very hard to catch up. If you're going to do this, the best time is to claim one circle in one region before the first half of 2027.

What would mean you got it wrong: ① KOCA grows quickly in Hong Kong and Malaysia, which means it can replicate itself and the regional window left for others is smaller than you think. ② Android daily installs keep falling below two hundred and the Taiwan chart rank keeps slipping, which means the Taiwan photocard market is itself small, and doing it in another small market hits the same ceiling. ③ Sellers return en masse to the groups because of fees and account bans, which means sellers don't buy the "platform guarantee" idea and the formula itself needs to change.

Data and sources

• App Store (Taiwan) rating count, average score, daily new ratings, Shopping free chart rank (10-28 to 11-06); latest App Store Taiwan reviews (September to 11-04).
• Google Play installs, 7-day growth, Taiwan Shopping free chart rank (10-29 to 11-06).
• Threads @koca.universe: 2025-04-23 FAQ post (11-06 reads, 64k views), monthly ranking, ban notices, Planet Captain recruitment post; posts from users such as @jminj_0401.
• Official IG: a post on the 4-month milestone since launch, and a post on the East District pop-up.
• Fengchuan Media 2026-10-24, NOWnews 2026-11-03 coverage (reposted on Yahoo and LINE TODAY).
• Official sites koca.app and koca.shop; LinkedIn company page; competitors' store pages and websites; the Dcard fan board and photocard trading discussions on Reddit.
• Where it says "inferred" or "unverified," it means inference or unchecked.