# The CEX User-Acquisition Funnel in Korea: Install, Sign-up, First Deposit

*ium Research — Tobi, Senior Analyst — June 13, 2026*
*Canonical: https://iumlabs.io/blog/korea-cex-user-acquisition-funnel-kol-naver-paid-ads*

## Key Takeaways

- Korea is one of the deepest retail crypto markets on earth: 9.7 million verified, trade-eligible users by the end of 2024, equal to roughly one in five Koreans (Source: [FSC/KoFIU semi-annual VASP survey](https://www.fsc.go.kr/po010101/84618)).
- The market is winner-take-most. Upbit alone held 71.6 percent of domestic trading volume in the first half of 2025, and Upbit plus Bithumb account for about 90 percent of all users (Source: [FSC/KoFIU VASP survey data](https://www.fsc.go.kr/po010101/84618)).
- The funnel does not break at the app store. It breaks at the deposit, where every won must move through a real-name bank account at the exchange's single partner bank (Source: [FSC/KoFIU semi-annual VASP survey](https://www.fsc.go.kr/po010101/84618); Paybis).
- Discovery still runs through Naver, which holds roughly 42 percent of Korean search and far more of the high-intent informational queries (Source: [Statcounter Global Stats, Korea](https://gs.statcounter.com/search-engine-market-share/all/south-korea), May 2026).

In most markets a growth team can hide behind installs and sign-ups. In Korea you cannot. A Korean crypto exchange play has one metric that pays rent: the funded account. An install is a number a vendor sells you; a sign-up is a phone-verified email. Neither moves revenue. Trading fees, the entire business model of a Korean exchange, begin only after a real person links a real-name bank account and pushes won across it. Everything upstream of that deposit is vanity. This report walks the funnel stage by stage and explains why each runs on a different Korea-native channel.

## 1. The Funnel and the Korean Numbers

By year-end 2024 Korea had 9.7 million verified users eligible to trade digital assets, up 25 percent in six months, meaning roughly one in five Koreans holds an exchange account. Won deposits at domestic exchanges more than doubled to 10.7 trillion won, and daily average volume reached 7.3 trillion won. The base skews young: investors in their 30s are the largest cohort at 29 percent, followed by 40s at 27 percent.

> Embedded post: https://twitter.com/build_on_bob/status/2013253580598268363

| Item | Value |
|---|---|
| Age 20s & under | 19 |
| Age 30s | 29 |
| Age 40s | 27 |
| Age 50s | 18 |
| Age 60s+ | 7 |

*Share of Korean verified crypto users by age cohort, end-2024. The 30s lead. Source: FSC / FIU, via KED Global*

This is a deep, fee-generating market, but a brutally concentrated one. The funnel does not just need to convert; it must convert against incumbents who already own the trust layer. The age skew is itself an operating instruction. A funnel aimed at the 30s and 40s, who together make up 56 percent of verified users, is a funnel aimed at where the deposits actually sit, not at the youngest cohort that indexes high on social reach but low on funded balances. Channel choice should follow the money in the book, not the noise on the timeline.

> **9.7 million verified users** — Roughly one in five Koreans holds a trade-eligible exchange account, up 25% in six months to end-2024 (Source: [FSC/KoFIU semi-annual VASP survey](https://www.fsc.go.kr/po010101/84618))

> **71.6%** — Upbit's share of Korean crypto trading volume, H1 2025 (Source: [FSC/KoFIU VASP survey data](https://www.fsc.go.kr/po010101/84618))

## 2. Where Korea Differs: the Deposit Is the Wall

In a typical Western funnel the painful step is sign-up. In Korea sign-up is trivial and the deposit is the wall. To accept won, a licensed exchange must hold a real-name account agreement with a single domestic bank, and every deposit and withdrawal must flow through an account in the user's exact legal name at that bank. Upbit pairs with K-Bank, Bithumb with KB Kookmin, Coinone with Kakao Bank, Korbit with Shinhan. As of early 2026 only five exchanges hold such partnerships at all.

The single-bank rule is not a detail; it is the structure of the whole market. It means each exchange's reachable audience is partly a function of which bank's customers it can plausibly convert, and it explains why trust and incumbency compound. The table makes the concentration concrete, pairing each licensed exchange with its bank rail and the user share the post documents.

| Exchange | Real-name bank partner | Share of users |
| --- | --- | --- |
| Upbit | K-Bank | 53% |
| Bithumb | KB Kookmin | 37% |
| Coinone | Kakao Bank | under 1M users, split three ways |
| Korbit | Shinhan | under 1M users, split three ways |
| Gopax | (licensed) | under 1M users, split three ways |

> Embedded video: https://www.youtube.com/watch?v=y5gnkr3fQUg

The friction is unforgiving. When Upbit switched on mandatory verification, users could not transact above 1 million won without completing KYC, and won balances in unverified accounts were returned outright. The system rejects foreign accounts, joint accounts, and name mismatches with no human review. A new user must, in one sitting, verify identity, hold or open an account at the correct bank, and pass the name-match check before a single won lands. That is where installs go to die.

*[Figure: Source: KED Global, "S.Korea's crypto market doubles to $77.5 bn with 20% of population trading"]*

This Korean walkthrough makes the friction visible: it has to teach viewers that you cannot deposit won directly to a global exchange and must instead route coins through Upbit, complete two-channel authentication, and wait on a transfer. Each added step is a place the funnel leaks.

> Embedded video: https://www.youtube.com/watch?v=AtRj5BD0bUg

## 3. Trust Gates the Conversion

The deposit is not only procedural friction; it is the most expensive trust decision a Korean retail user makes. Linking a real-name bank account hands an exchange a verified identity tied directly to a bank ledger. Koreans do this readily for Upbit and Bithumb because those brands are effectively household infrastructure. They do not do it for an unknown brand on the strength of a banner ad.

*[Figure: Kaiko, market share of major Korean exchanges shows Upbit dominance through 2025]*

The concentration data proves it. Of roughly 10.2 million users, 53 percent sit on Upbit and 37 percent on Bithumb, leaving the other three licensed exchanges to split under one million users. Trust, not features or fees, sorts that distribution.

| Item | Value |
|---|---|
| Upbit | 53 |
| Bithumb | 37 |
| Other three | 10 |

*Share of roughly 10.2 million Korean exchange users; trust, not fees, sorts the distribution. Source: FSC / FIU, via KED Global*

> **"The domestic crypto market has effectively solidified into an Upbit monopoly."**
>, Seoul Kyungjae, via Cryptonews

That should reframe how an operator thinks about spend. You are not buying clicks. You are buying enough credibility for a stranger to attach their bank identity to your brand. Paid media cannot manufacture that alone. It is earned upstream, by voices the user already trusts, and only then scaled. This is also why a late-entrant exchange cannot simply outspend the incumbents into relevance; the 53 and 37 percent shares are not a media-budget gap, they are a trust gap, and trust is bought in a different currency than impressions.

## 4. What This Means for Operators

Run the funnel as three distinct jobs on three Korea-native channels, and price the whole thing on one number: cost per funded account.

Discovery runs on Naver, not Google. Naver holds roughly 42 percent of Korean search and a far larger slice of the long-form informational queries where a cautious first-time depositor researches an exchange, because so much Korean-language content lives inside Naver Blog and Cafe. If your brand is absent from Naver's index and communities, the consideration stage never happens. For the deeper mechanics of why Korean discovery behaves this way, see [the Naver effect](/blog/the-naver-effect-search-dominance-korean-crypto-discovery).

Trust runs on KOLs. A Korean crypto creator with a years-built audience does the one thing a banner cannot: vouch credibly enough that a viewer will complete real-name KYC and deposit. This is the conversion engine, not an awareness line item. How that creator layer is structured and priced is its own discipline, covered in our [KOL marketing landscape](/blog/korea-kol-marketing-landscape-2026-guide).

Scale runs on paid. Once Naver is seeding discovery and KOLs are carrying trust, paid media amplifies a funnel that already converts. Paid is the multiplier, never the foundation.

| Item | Value |
|---|---|
| Naver | 42 |
| Google | 48 |
| Bing | 6 |

*South Korea search engine share, May 2026. Naver still owns the high-intent informational layer. Source: Statcounter*

## 5. What Breaks It

The fastest way to burn a Korean exchange budget is to optimize to the wrong metric. Buy installs and you get installs, sit at the deposit wall, and report a CAC that looks fine next to an empty revenue line.

Three failure modes recur. Google-first SEO and Western paid playbooks that ignore Naver and miss the user during research. Treating KOLs as a cheap reach buy rather than the mechanism that unlocks the deposit, which yields views and no funded accounts. And optimizing to sign-ups because they are easy to measure, while the real-name bank link, the actual bottleneck, goes unmanaged. If your dashboard cannot show cost per funded account by channel, you are flying blind through the only stage that matters.

There is a structural counter-case too. Because Upbit and Bithumb already hold about 90 percent of users, a new entrant should be honest that the realistic prize is rarely dethroning the duopoly; it is capturing a defensible slice of the deposit flow the incumbents leave on the table, or converting a specific cohort the majors under-serve. A funnel built on the fantasy of mass switching will overspend at the trust gate and still stall at the bank link. A funnel built on a narrow, winnable wedge, priced on cost per funded account, can clear it.

## 6. How We Run It

ium Labs builds the Korean exchange funnel backward from the funded account. We map the deposit friction for your specific bank partner and onboarding flow, then assign each stage to the channel that moves it: Naver search and community presence for discovery, vetted Korean crypto KOLs for the trust gate that unlocks real-name KYC, and paid media to scale only once conversion is proven. Every channel reports to one metric, cost per funded account, so spend follows deposits, not vanity volume.

If you are bringing an exchange or trading product into Korea and need a funnel measured in funded accounts, see how we run [exchange marketing](/services/exchange-marketing) or [get in touch](/contact).

## Sources

- [KED Global: S.Korea's crypto market doubles to $77.5 bn with 20% of population trading](https://www.kedglobal.com/cryptocurrencies/newsView/ked202505200007)
- [Cryptonews / Yahoo Finance: Upbit Corners 72% of S Korean Crypto Market](https://finance.yahoo.com/news/upbit-corners-72-korean-crypto-233000204.html)
- [Maeil Business / Pulse: Upbit to mandate users to verify identity for crypto trade](https://pulse.mk.co.kr/news/english/10043646)
- [Paybis: Best Crypto Exchanges in South Korea 2026](https://paybis.com/blog/best-crypto-exchanges-korea/)
- [Statcounter: Search Engine Market Share Republic of Korea](https://gs.statcounter.com/search-engine-market-share/all/south-korea)
- [YouTube: Binance sign-up and Upbit deposit guide (2026)](https://www.youtube.com/watch?v=AtRj5BD0bUg)
- [YouTube, Upbit Tutorial 2025 Account Setup Security Transfers Withdrawals](https://www.youtube.com/watch?v=y5gnkr3fQUg)
- [Bitcoin.com, Upbit Towers Over Rivals as South Korea Primary Liquidity Hub Kaiko](https://news.bitcoin.com/upbit-towers-over-rivals-as-south-koreas-primary-crypto-liquidity-hub-research-finds/)

*This report reflects ium Labs' operating view and is intended for general information, not investment advice.*