There are three ways to sell into Indonesia: cross-border marketplace selling, local selling via an importer of record (IOR), and your own PT PMA. Cross-border is cheap to start but restricted categories and slow delivery kill conversion. An IOR adds import duty and PPN to landed cost but unlocks local delivery speed, COD and regulated categories. A PT PMA maximises control and margin but carries fixed costs and an IDR 10 billion investment plan per KBLI code. Most brands should climb the ladder in that order.
The question is rarely "should we sell in Indonesia?" but "through which structure?" — and the honest answer changes as your volume grows. Each model trades margin against commitment differently, and picking the wrong rung of the ladder either starves your margin or ties up capital before the market has voted. Here is the truth-table.
What does each model actually involve?
Cross-border means fulfilling Indonesian marketplace orders from a Singapore or regional warehouse through a platform's cross-border scheme. No Indonesian entity, no import licences — but category restrictions apply (regulated products such as food, supplements and cosmetics generally cannot ship this way at scale), low-value import concessions are narrow and repeatedly tightened, and delivery takes days longer than local sellers. In a market where next-day Java delivery is normal, that lag shows up directly in conversion.
Importer of record (IOR) means a licensed local importer brings your goods in under their licences, you hold stock in a local 3PL, and you sell as a local seller. You absorb import duty and PPN into landed cost, but you get local delivery speeds, cash-on-delivery access, participation in campaigns, and legal routes for BPOM-regulated products. How this works in detail is covered in our importer of record guide.
Own PT PMA means your own foreign-owned Indonesian company with its own import rights. Full control of registrations, contracts and margin — and full fixed costs: incorporation, directors, accounting, tax compliance and an investment plan of IDR 10 billion per KBLI business classification code. The set-up path is described in our PT PMA guide.
How do the economics compare line by line?
| P&L line | Cross-border | IOR local selling | Own PT PMA |
|---|---|---|---|
| Set-up cost | Minimal | Low (IOR onboarding) | High (entity, capital plan) |
| Import duty + PPN | Buyer-side / scheme-dependent | In landed cost | In landed cost |
| IOR / handling fees | Platform cross-border fees | Per-shipment IOR margin | None (own licences) |
| Marketplace take rate | Higher cross-border rates | Local rates (Shopee 2.5-17.4%) | Local rates |
| Delivery speed (Java) | 5-10+ days | 1-3 days | 1-3 days |
| COD access | No | Yes | Yes |
| Regulated categories (BPOM) | Largely blocked | Yes, via licence holder | Yes, own registrations |
| Fixed monthly overhead | Near zero | Low | Significant |
| Control of registrations/data | n/a | Shared (keep registrations neutral) | Full |
Two dynamics deserve emphasis. First, the take-rate squeeze: with Shopee at 2.5-17.4% plus a Rp1,250 per-order fee, Tokopedia at 5-15.8% and Lazada up to 18.24% per 2025 rate cards, every other cost line matters more than it did in 2022 — our margin planning guide works the full waterfall. Second, conversion: cross-border's slower delivery does not just lose the impatient shopper, it excludes you from COD, which remains significant outside Java. A thinner-margin local model that converts at multiples of cross-border rates usually wins on absolute contribution.
When should you switch models?
Useful decision thresholds, expressed in monthly GMV terms rather than hard rules:
- Testing demand (roughly below S$10k/month): cross-border only if your category permits it and delivery expectations are modest; otherwise go straight to IOR — regulated products have no legitimate cross-border shortcut anyway.
- Proving and growing (roughly S$10k-100k/month): IOR-based local selling is the sweet spot: real conversion data, campaign access, no fixed entity burden. Keep BPOM registrations with a neutral licence holder, not a distributor.
- Committed scale (sustained six figures monthly, or offline distribution): the IOR's per-shipment margin and shared control start costing more than a PT PMA's fixed overhead. Incorporate, migrate registrations to your own entity, and take your own import licences.
The pattern to avoid is the inverted one: incorporating first "to do it properly", then discovering demand does not justify the overhead. The entity should be a consequence of evidence, not a prerequisite for it.
How does the migration path map to the MRA grant?
The three-stage ladder maps neatly onto the MRA grant's three pillars, each supportable at up to 70%, subject to Enterprise Singapore's approval. Stage one: validate demand through an IOR-shipped live pilot — promotion and campaign costs sit under Overseas Market Promotion, which is how our 90-day validation programme comes down from S$28,000 to from S$8,400. Stage two: distributor identification and in-market business development under Overseas Business Development. Stage three: entity incorporation, licensing, tax structure planning and agreement drafting under Overseas Market Set-up — the scope of our market set-up service. Sequenced properly, the grant defrays each rung of the ladder as you climb it.
Cross-border tests curiosity, an IOR tests the market, a PT PMA banks the win. Climb in that order — the entity should follow the evidence, not precede it.
Key takeaways
- Cross-border is cheapest to start but slow delivery, no COD and category restrictions cap its ceiling.
- IOR-based local selling is the workhorse middle rung: local conversion economics without entity fixed costs.
- A PT PMA pays off at sustained scale, but carries an IDR 10 billion investment plan per KBLI code.
- Rising take rates (up to 17-18% on major platforms) make every other margin line more consequential.
- The validate → IOR → entity path maps onto the MRA's three pillars at up to 70% support, subject to Enterprise Singapore's approval.
Frequently asked questions
Can I sell to Indonesia from Singapore without a local entity?
Yes, two ways: cross-border marketplace schemes fulfilled from outside Indonesia, or local selling through an importer of record who brings goods in under their licences while you hold stock at a local 3PL. The IOR route unlocks local delivery speed, COD and regulated categories, which cross-border cannot offer at scale.
When does a PT PMA make financial sense?
Generally at sustained six-figure monthly GMV or when offline distribution begins. At that point the IOR's per-shipment margin and shared control cost more than the entity's fixed overhead. Remember the commitment: a PT PMA requires an investment plan of IDR 10 billion per KBLI business classification code, plus ongoing compliance.
Why does cross-border selling convert poorly in Indonesia?
Delivery and payment. Local sellers reach most of Java in one to three days while cross-border parcels take five to ten or more, and cross-border listings cannot offer cash on delivery, which remains important outside Java. Shoppers comparing identical products systematically choose the faster, COD-capable local listing.
Model your Indonesia P&L before you commit
Book a complimentary 1-hour discovery session. We'll work your landed-cost and margin numbers across all three models and recommend the right rung to start on.
Book a free discovery sessionRelated reading: Importer of record in Indonesia · Marketplace fees and margin planning · Setting up a PT PMA
Take rates per Databoks/Katadata (2025); duties, taxes and thresholds change — verify current rates before finalising pricing. Not tax advice.