The platform split, and why it matters
Shopee leads Indonesian ecommerce, and TikTok Shop became the clear second engine after ByteDance took control of Tokopedia in December 2023 and merged its Indonesian shopping operation into the platform. But treating this as a simple ranking misses the point. The two platforms reward completely different behaviour, and a strategy tuned for one performs poorly on the other.
Shopee, search led
The buyer arrives knowing roughly what they want. Availability, price position, review volume and delivery speed determine whether a listing wins.
TikTok Shop, discovery led
The buyer did not set out to buy. Content velocity, live commerce and creator economics determine whether a product is encountered at all.
Category share is two separate wins
Because the journeys differ, share on one platform tells you almost nothing about share on the other. Measuring one while flying blind on the other is not a viable operating model.
Live commerce is not a side channel
Video commerce grew from a marginal share of Indonesian online GMV in 2022 to roughly a fifth by 2025. That is a structural change in how products are discovered and sold, not a format experiment. Conversion behaviour differs from catalogue listings, and the operational requirements are different too: live commerce needs scheduling, hosts, inventory synchronisation and a content cadence rather than a campaign burst. For a foreign brand entering now, this is a capability decision to make before launch, not after.
Entering Indonesia is not about migrating platforms
It is about rebuilding the decision-making system. The functions look familiar. What sits behind each one, and how the pieces connect, does not.
| Function | What it maps to in Indonesia | What is different |
|---|---|---|
| Messaging | WhatsApp, used monthly by nine in ten Indonesians | Not a chat app in the marketing sense. It is a CRM and commerce channel, yet it sells almost no formal advertising, so it is absent from every media plan |
| Video discovery | TikTok at 180 million ad reach, YouTube at 151 million | TikTok reaches 88.9 percent of adults 18 and over and holds the highest daily time at 1 hour 53 minutes |
| Research and consideration | Google, with social running almost level | Search accounts for 38.3 percent of brand discovery, social ads 37.3 percent. Functionally tied |
| Community | WhatsApp groups and comment sections | Largely closed. Standard listening tools see very little of it |
| Commerce | Shopee and TikTok Shop | Two engines with opposite logic. One is search led, the other discovery led |
| Payment | QRIS, a single national QR rail | The inverse of fragmentation. One acceptance point reaches essentially every wallet and bank app |
Platform reach: DataReportal Digital 2026. Brand discovery split: We Are Social and Meltwater Digital 2026. Marketplace and payment structure: Arfadia Digital Marketing Benchmark Indonesia 2026.
The conclusion is the opposite of China
In China, one ecosystem covers many scenarios and purchase closes inside it. In Indonesia the platforms stay separate, with two commerce engines that reward opposite behaviour, but they share a single payment rail. A brand that arrives expecting either a Western stack or a Chinese super app will misread both halves.
How an Indonesian buyer actually decides
Five stages, each dominated by a different platform. The stage most foreign brands underinvest in is the third.
TikTok and Instagram
Discovery is video first. TikTok holds the highest daily time of any platform at 1 hour 53 minutes.
Social platforms
Six in ten Indonesians use social media as their primary channel for researching brands.
Google, and increasingly AI assistants
Search still leads brand discovery at 38.3 percent. More than a third of Indonesians now use ChatGPT every month.
Shopee and TikTok Shop
Live commerce converts at 5 to 12 percent, roughly three times catalogue listings. Sixty percent of video commerce buyers purchase during a live session.
WhatsApp and comment sections
Social comments and posts account for 32.6 percent of brand discovery, which feeds the next buyer's attention stage.
Daily time and platform reach: DataReportal Digital 2026. Brand research behaviour and discovery split: We Are Social and Meltwater Digital 2026. Live commerce conversion: Arfadia Digital Marketing Benchmark Indonesia 2026, citing e-Conomy SEA 2025.
The third stage is where foreign brands lose the most ground. A brand can win attention with paid video and still lose the sale, because the buyer leaves the platform to verify the claim and finds nothing. Search and AI visibility is not a separate channel from social. It is the step that decides whether social spend converts.
QRIS, one rail rather than many wallets
This is where most international merchants build the wrong integration. Indonesia does not have a dominant wallet that you integrate with. It has a dominant rail. QRIS, the unified QR standard set by Bank Indonesia, sits beneath GoPay, OVO, DANA, ShopeePay and most bank applications. A single QRIS acceptance point reaches essentially the entire digitally active population, which means integrating wallets individually is redundant work.
One code, all wallets
The customer opens whichever app they already use and scans the same code. No wall of wallet logos required.
Cross border reach
QRIS interoperability with several regional markets means the same acceptance point also serves visiting consumers from those countries.
Aggregator route is standard
Foreign operators typically enter through licensed local aggregators rather than seeking a direct central bank licence, because foreign ownership in payment service providers is constrained and domestic parties must retain voting control.
Where foreign brands lose time
Running Indonesia entirely from abroad
Local warehousing shortens delivery time, and delivery time feeds marketplace ranking. A product shipped from overseas on every order tends to lose visibility to a locally stocked competitor.
Underestimating language expectations
Support answered in English after a long delay creates a first impression that is difficult to reverse. Native language response is an expectation, not a differentiator.
Treating both platforms as one channel
A single content and pricing strategy applied to both engines typically underperforms on both.
Integrating wallets one by one
Redundant engineering that produces a worse checkout than a single QRIS acceptance point.
Where Arfadia fits
Arfadia does not handle logistics, warehousing or payment licensing. What Arfadia does is the demand side: making a product findable on the platform where its category actually converts, producing the content velocity that discovery led platforms require, and connecting marketplace presence to search and AI visibility so the three reinforce each other rather than running as separate budgets.
Commission structures, seller programmes and cross border eligibility on Indonesian marketplaces change often and vary by category. Figures quoted in secondary sources go stale quickly. Confirm current terms directly with the platform before building a business case.
- Bank Indonesia QRIS standard and published adoption data
- Platform seller documentation for Shopee and TikTok Shop
- Reported ecommerce market tracking for Indonesia
- Bank Indonesia regulation on payment service provider ownership
Frequently Asked Questions
Do we need an Indonesian entity to sell on marketplaces?
Shopee or TikTok Shop, which should we prioritise?
What is QRIS and do we need it?
Can we get a payment licence directly?
How important is live commerce really?
Does Arfadia handle marketplace operations?
Listed everywhere, found nowhere?
Marketplace presence and search visibility are different problems. We work on the second.
Talk to Arfadia See client portfolio