DEV Community

Jim Yeh
Jim Yeh

Posted on

I Ran Cross-Border E-Commerce from China for 10 Years. Here's What Actually Works in Southeast Asia in 2026

I sold my first batch of products to Indonesia from a cramped office in Shenzhen in 2016. We printed labels late into the night, boxed orders by hand, and watched the tracking numbers inch across courier dashboards while hoping payments would clear. That first month I learned two things: the market was enormous, and the playbook I needed did not exist.

Over the next decade I built systems, learned messy lessons the hard way, and scaled from a handful of daily sales to consistent, predictable revenue across multiple Southeast Asian countries. This article captures what actually works in Southeast Asia in 2026 — the tactics I used, proven by real cases inside the playbook I am releasing.

1) Stop guessing platforms — pick the one that matches your product and margins

When I started, I treated Shopee, Lazada, Tokopedia and TikTok Shop as interchangeable channels. After testing the same SKU across three platforms for four months, the truth became obvious: each platform attracts different buyer intent and fee structures. One product I sold (a travel power bank) performed poorly on Lazada because price-sensitive customers were comparing against local sellers. The same SKU on Shopee, with a targeted voucher and bundle, sold three times as many units with a higher conversion rate.

What I did differently: I mapped unit economics (COGS + shipping + platform fee + ad spend) per platform before scaling. For example, a $5 product on Tokopedia with a 6% platform fee and 15% voucher budget still left room for profit if shipping could be consolidated. That decision saved us from pouring ad spend into a low-margin channel.

Actionable takeaway: create a simple spreadsheet that calculates "net margin per sale" for each platform. Test with clear sample runs (200–500 units) and double down on the channel that gives consistent ROAS above your break-even threshold.

2) Localized creatives and language beat generic listings every time

In 2018 we launched a health supplement across three SEA markets. Initially we only translated the title and called it a day. Sales were slow. When we invested in localized creatives — video thumbnails showing local currency, testimonials from local users (translated), and short captions in local dialect — conversion doubled.

Case study: For Indonesia, an A/B test of banner text in Bahasa with a short CTA ("Order sekarang") delivered a 45% uplift vs English-only creatives. For the Philippines we used localized imagery that referenced local shipping holidays and saw similar gains.

Actionable takeaway: localize more than language. Use visuals, social proof, and value propositions aligned to local habits (e.g., cash-on-delivery trust cues in markets where COD is common).

3) Build logistics playbooks around return friction and shipping cost optimization

Shipping in SEA is varied: islands, poor last-mile infrastructure, and high returns for certain categories. Early on we lost margin from a product with high return rates because our return window was generous but unclear. We reworked packaging, added clear size charts, and partnered with a regional fulfillment provider for consolidated shipping hubs.

Case study: We switched from single-package international shipments to regional consolidation hubs in Singapore and Malaysia. That change cut per-order shipping costs by ~18% and decreased delivery times for major markets.

Actionable takeaway: track return reasons and iterate packaging and listings to reduce preventable returns. Negotiate consolidation and volume discounts with logistics partners once you pass the first 500–1,000 monthly orders.

4) Use pricing psychology and local payment patterns, not global assumptions

Many sellers assume global buyers will accept the same checkout experience. In reality, payment preferences vary wildly. Vietnam and Indonesia still show high COD usage in certain categories, while Malaysia and Singapore are card-heavy and accept cross-border payments smoothly.

Experiment: We introduced a small early-bird discount (3–5%) and a time-limited voucher during local festivals. The result: spike in day-one sales and improved placement in search results across marketplaces (some platforms favor early velocity). We learned to pre-schedule discounts around local shopping dates rather than global Western holidays.

Actionable takeaway: segment your launch calendar by local events (e.g., Harbolnas-like sales, Ramadan, Eid-adjacent promotions) and provide payment options that match buyer comfort (COD, installment plans where supported, local e-wallets).

5) Document repeatable systems — then automate the low-value tasks

Most of our early operational failures were avoidable: inconsistent CSV uploads, misplaced inventory counts, and ad campaigns with duplicated audiences. We documented every step — SKU naming conventions, CSV templates per platform, and a weekly checklist for inventory sync. Then we automated: template-based spreadsheet generation, scheduled inventory pushes, and auto-bidding rules in ads for steady ROAS targets.

Case study: After implementing template-driven uploads and an automated monitoring script that flagged inventory variance >5%, manual errors dropped by 72% and stockouts became rare.

Actionable takeaway: before hiring, build "what-to-do" playbooks for the first five hires. If a task can be done by a template or scheduled script, automate it.

Real numbers and how I enforced discipline

I believe numbers keep you honest. In 2019 a single product line produced $6,000 per month GMV with manual processes. After systemizing pricing, creatives, and logistics, that same line scaled to $28,000 monthly GMV within 10 months without adding headcount. The difference: discipline in testing and a focus on unit economics.

How I structured experiments:

  • Hypothesis (e.g., localized CTA increases conversion by 20%)
  • Test size (minimum 200 units or 2 weeks)
  • Winner criteria (statistical uplift + stable margin)
  • Rollout plan (bundle, creative, and listing updates)

Why early launches fail and what to do instead

New sellers often fail by trying to optimize everything at once: price, creative, audience, shipping. The right approach is sequential: stabilize unit economics, then optimize traffic efficiency, then scale logistics.

Start with a narrow experiment: one SKU, one market, one ad channel. Only once repeatable results exist should you diversify. Diversity before profitability creates noise and hidden losses.

What this playbook contains (short preview)

Inside my guide I lay out the chapters I wish I had ten years ago:

  • Chapter 1: Market Overview — country-level buyer behavior and platform fit
  • Chapter 2: Platform Selection — which channels make sense for which SKUs
  • Chapter 3: Product Selection & Sourcing — a $1,000/day framework with supplier scripts
  • Chapter 4: Logistics & Fulfillment — consolidation, partner negotiation, and packaging
  • Chapter 5: Payments & Financial Management — local payment rails and tax basics
  • Chapter 6: Localization & Marketing — creatives, translations, and social proof
  • Chapter 7: Operations & Optimization — checklists and automation templates
  • Chapter 8: Scaling & Exit Strategies — metrics to watch and acquisition preparedness

If you want the full playbook with all 8 chapters and 15 templates, I put everything I learned into a $4.99 guide here: https://jimyeh.gumroad.com/l/edxwso

Closing note

Selling into Southeast Asia from China is not a mythical shortcut — it is operational craft. If you can get the unit economics right, localize with respect, and build repeatable systems, you can build a margin-first business. I hope these lessons help you skip a few painful months of trial and error.

— Jim Yeh

Top comments (0)