← 返回所有筆記 ← Back to all notes

創辦人筆記 Founder note

我喺大集團學到嘅 multi-currency 教訓 — 點解中小品牌唔應該行同一條路 What I learned about multi-currency at an HK$8B group — and why HK SMEs shouldn't follow that path

我喺一間 Euronext 上市嘅奢侈時裝集團做亞洲區 BI 主管嘅時候，我哋有一個 quarter 嘅 OKR 係：「Reconcile multi-currency settlement across 5 markets to within 0.5% variance.」

聽落好 boring。但呢個係一個年收入超過 80 億港元嘅集團，5 個市場（HK、AU、NZ、SG、CN），4 隻幣種（HKD, AUD, NZD, SGD），3 個 PSP（Adyen, Stripe, local processors），2 個 ERP layer。每個月幾百萬筆 transaction。0.5% variance 喺 80 億港元收入 scale 係 4,000 萬港元 — 實質金額大過好多香港品牌嘅一整年生意。

呢場仗值得打。但我見過太多香港中小品牌，年做 HK$3M-30M，俾人灌輸要解決同一個問題。99% 係 over-engineering。

我喺大集團學到嘅嘢

1. Multi-currency 嘅成本，9 成係 ops 唔係 tech。

買套 multi-currency-capable 嘅 platform 好易（Shopify Markets, Stripe, Adyen 全部做到）。難嘅係：每日 reconcile bank statement 對唔對得上 PSP settlement 對唔對得上 ERP entry。一個 transaction 喺 Adyen 用 SGD record，settle 落你 SGD bank account，但你嘅 HK 總公司用 HKD report，月底要做 conversion + accrual。呢啲全部係人手做，唔係 software 做。

我哋集團有 4 個 finance ops 員工專做 reconciliation。對中小品牌嚟講，呢 4 個人就係你自己。

2. 大部份品牌唔需要 multi-currency settlement，需要嘅係 multi-currency display。

呢兩個係好唔同嘅嘢：

  • Multi-currency display：客戶喺 product page 睇到 HKD 或 USD，但 checkout 時統一用一隻幣（例如 HKD）扣錢。技術好簡單，rate 用每日 ECB rate 就夠。客戶買嘅體驗好過全英文兼純 USD。
  • Multi-currency settlement：客戶用美金扣，你美金收，月底再 convert 返 HKD 入帳。呢個會引發上面講嘅 ops nightmare。

90% 中小品牌真正需要嘅係前者。但 sales 通常 push 你買後者。

3. 對小品牌嚟講，「賺嘅 FX margin」往往細過「俾錢嘅 FX 服務費」。

舉個例：你係香港品牌，幫海外客戶 (主要美國) display USD price。如果你用 Shopify Markets settlement in USD，每筆 transaction 收 1.5% currency conversion fee。你以為呢 1.5% 係幫客戶轉去 USD？唔係 — 係 Shopify 賺嘅 FX margin。

如果你純粹 display USD（用每日 reference rate）但實際 charge HKD：客戶體驗冇分別，你慳返 1.5%。對年做 HK$1M 嘅店，呢個係 HK$15,000/年，等於請半年 PT 設計師。

4. Currency switching 嘅 conversion lift 細過你以為。

我哋集團做過 A/B test：show local currency vs USD 對 Asian markets 嘅 conversion lift 大約 8-15%（睇 market）。但呢個 lift 嘅 baseline 係：你個 storefront 已經有英文以外嘅內容、有 trust signal、有本地支付方式。如果你 storefront 仲係純英文 + 只收 Visa/Master，加 currency switcher 唔會幫到你。

優先 sequence 永遠係：本地語言 → 本地支付 → 本地 currency display → multi-currency settlement。

中小品牌應該點做

我寫 Kongmerce 嘅時候，呢條經驗直接 shape 咗 product decision：

  • MVP： 單幣（HKD）, FPS / 銀行轉帳 / 自取 / 貨到付款 — 因為 80% HK SME 嘅 80% 客戶 fall in 呢個 case
  • Phase 1： Stripe Connect 卡支付（HKD settlement）— 服務海外客戶但仍然 settle HKD
  • Phase 2： Multi-currency display （無 settlement）— 用每日 reference rate，display USD/CNY/GBP，但 charge HKD
  • 未必會做： Multi-currency settlement — 因為 99% target customer 唔需要，但會 add 好多 ops 複雜度

呢個 sequencing 係用 15 年大集團 lessons 倒推番畀小品牌。Enterprise tooling 嘅縮水版，唔係 SME tooling。

我嘅 take

如果你係小品牌，下次有人 sell 你「multi-currency support」，問清楚：

  1. 係 display 定 settlement？
  2. 你嘅客戶真係要扣本地幣，定睇本地幣已經夠？
  3. 多收嘅 FX fee 對你 P&L 影響有幾大？
  4. 月底 reconcile 你打算自己做、定請人做？

99% case 你會發現你需要嘅係簡單版。慳返嘅錢同時間，投落 marketing 同產品上面。

想傾下你個 case 應該點？WhatsApp 我。

When I was Head of Asia BI at a Euronext-listed luxury fashion group, one of my quarterly OKRs was: “Reconcile multi-currency settlement across 5 markets to within 0.5% variance.”

Sounds boring. But the group did HK$8B+ in revenue across 5 markets (HK, AU, NZ, SG, CN), 4 currencies (HKD, AUD, NZD, SGD), 3 PSPs (Adyen, Stripe, local processors), and 2 ERP layers. Hundreds of thousands of transactions a month. 0.5% variance against HK$8B is HK$40M — more than the annual revenue of many HK brands.

That fight was worth fighting at that scale. But I’ve watched too many HK SMEs doing HK$3M-30M/year get sold the same problem. 99% of the time it’s over-engineering.

What I learned at scale

1. Multi-currency cost is 90% ops, 10% tech.

Buying a multi-currency-capable platform is easy (Shopify Markets, Stripe, Adyen all do it). The hard part is reconciling daily: bank statement → PSP settlement → ERP entry. A transaction recorded in SGD by Adyen, settling into your SGD bank account, but reported in HKD by the HK parent, requires conversion + accrual at month-end. All of that is humans, not software.

We had 4 finance ops headcount dedicated to reconciliation. For an SME, those 4 people are you.

2. Most brands don’t need multi-currency settlement; they need multi-currency display.

These are very different things:

  • Multi-currency display: customers see HKD or USD on the product page, but checkout charges in a single currency (e.g., HKD). Technically simple — daily ECB reference rate is enough. The customer experience is much better than English-only and USD-only.
  • Multi-currency settlement: customer is charged in USD, you receive USD, you convert to HKD at month-end. This triggers the ops nightmare above.

90% of SMEs only need the first. But sales teams typically push you to buy the second.

3. For small brands, the FX margin you “earn” is usually smaller than the FX fee you “pay” to the platform.

Example: you’re an HK brand showing USD prices to overseas (mainly US) customers. If you use Shopify Markets settlement in USD, every transaction pays a 1.5% currency conversion fee. You might assume the 1.5% goes to actually converting to USD for the customer. It doesn’t — that’s Shopify’s FX margin.

If you simply display USD (using a daily reference rate) but actually charge HKD: the customer experience is identical, and you save the 1.5%. For a shop doing HK$1M/year, that’s HK$15,000/year — equivalent to 6 months of a part-time designer.

4. Currency switching’s conversion lift is smaller than you think.

We A/B tested this at the group: showing local currency vs USD across Asian markets lifted conversion 8-15% depending on market. But that lift assumes you’ve already cleared the table stakes: localized content, trust signals, local payment methods. If your storefront is still English-only and accepts only Visa/Mastercard, adding a currency switcher won’t move the needle.

The right sequencing is always: local language → local payment → local currency display → multi-currency settlement.

What SMEs should actually do

This experience directly shaped Kongmerce’s roadmap:

  • MVP: Single currency (HKD), FPS / bank transfer / self-pickup / COD — because 80% of HK SMEs’ customers fall into this case
  • Phase 1: Stripe Connect cards (HKD settlement) — to serve overseas customers without leaving HKD reporting
  • Phase 2: Multi-currency display (no settlement) — daily reference rate, show USD/CNY/GBP, charge HKD
  • Probably never: Multi-currency settlement — 99% of target customers don’t need it, but it adds enormous ops complexity

This sequencing is enterprise lessons reverse-engineered for small brands. Not enterprise tooling, scaled down. Real SME tooling.

My take

If you’re a small brand and someone is selling you “multi-currency support”, ask:

  1. Is this display or settlement?
  2. Do customers actually need to be charged in their local currency, or is seeing the local currency enough?
  3. What’s the FX fee impact on your P&L?
  4. Who’s reconciling at month-end — you, or someone you’ll hire?

99% of the time you’ll find you need the simpler version. The money and time saved goes back into marketing and product — the things that actually grow the business.

Want to talk through your specific case? WhatsApp me.