Shopee or TikTok Shop: Decide by Category

Ask which Indonesian marketplace a foreign brand should prioritise and you will usually get a market share figure in reply. That answer is less useful than it looks, because the two leading platforms are not competing for the same moment in a buyer's day.

Shopee is where someone goes having already decided roughly what they want. TikTok Shop is where someone buys something they were not looking for. Those are different businesses wearing the same label, and a strategy tuned for one tends to underperform on the other.

This article sets out what the data actually says, why the market share question is the wrong one, how to read the Shopee and Tokopedia paradox, and how to decide by category rather than by ranking. Figures are cited to their sources throughout.

Why market share alone misleads

Two credible measurements of Indonesian marketplace dominance exist, and they disagree. Not because one is wrong, but because they measure different things.

The Shopee and Tokopedia paradox, and what to do with it

That last point deserves its own paragraph, because it inverts the usual conclusion.

If you rank by how often Indonesians open an app, Tokopedia looks like a minor player at 9.57 percent against Shopee's 53.22 percent. If you rank by transaction value, Tokopedia holds 23 percent against Shopee's 46 percent. Five and a half times the visits, roughly twice the value.

The implication is uncomfortable for anyone building a plan from a single chart. A brand selling low ticket consumables should weight toward frequency, because that is where its buyers are. A brand selling a two million rupiah appliance should not, because visit counts tell it almost nothing about where transactions of that size actually close.

This is also why "which platform is biggest" is a question worth refusing. Biggest by what, for whose basket size.

Two engines, opposite logic

Underneath the numbers sits a structural difference that matters more than share.

Shopee behaves as a search led environment. The buyer arrives with intent, types something, and compares. What wins a listing is availability, price position, review volume and delivery speed. It rewards catalogue discipline, competitive pricing and accumulated social proof.

TikTok Shop behaves as a discovery led environment. The buyer did not set out to buy anything. What determines whether a product is encountered at all is content velocity, live commerce activity and creator economics. It rewards production capacity and consistency rather than catalogue optimisation.

Live commerce is not a format experiment

The scale of the shift is worth stating with numbers, because it changes what a TikTok Shop presence costs to run.

Video commerce reached 2.6 billion transactions in Indonesia in 2025, up 90 percent year on year, with roughly 800,000 sellers, up 75 percent. It accounts for around 20 percent of online GMV, with average order value between Rp 70,000 and Rp 93,000. Conversion runs about three times catalogue listings, and 60 percent of video commerce buyers purchase during a live session rather than afterwards.

Those figures come from e-Conomy SEA 2025 as compiled in Arfadia's benchmark research. What they mean operationally is that live commerce needs scheduling, hosts, inventory synchronisation and a content cadence. It is a staffing decision, not a campaign line, and it has to be made before launch rather than after the first quarter underperforms.

Reviews are a barrier to entry, and they are asymmetric

One requirement separates the two engines more sharply than anything in a media plan, and new entrants consistently underestimate it.

Search led platforms rank partly on accumulated social proof. A listing with twenty to fifty reviews behaves very differently from one with none, because reviews feed both the algorithm and the buyer's decision. That threshold is a genuine cold start problem: you cannot buy your way past it quickly, and the first months of a Shopee presence are spent building something that competitors already have.

Discovery led platforms are less exposed to this. Content can carry a product that has almost no review history, because the buyer is being persuaded in the moment rather than comparing options. Ten to thirty reviews is a more typical working threshold, and a strong live session can convert around the gap.

The strategic implication runs against intuition. For a brand with no Indonesian sales history, the discovery led platform is often the faster route to first revenue, even though the search led platform is larger. And the revenue earned there produces the review base that makes the search led platform viable later. Sequencing the two, rather than choosing between them, is usually the better plan.

Deciding by category rather than by ranking

The useful question is not which platform is larger. It is whether your category is bought deliberately or encountered.

The mistake that costs the most

Running Indonesia entirely from abroad. Local warehousing shortens delivery time, and delivery time feeds marketplace ranking on search led platforms. A product shipped from overseas on every order tends to lose visibility to a locally stocked competitor even when the price is better.

Two more recur. Answering support in English after a long delay creates a first impression that is difficult to reverse, because native language response is an expectation rather than a differentiator. And applying a single content and pricing strategy to both engines typically underperforms on both, since each rewards behaviour the other does not.

What sits underneath both platforms

One structural fact simplifies the payment question considerably, and international merchants routinely build the wrong thing here.

Indonesia does not have a dominant wallet 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 engineering that produces a worse checkout.

Foreign operators typically reach it 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.

What a realistic first year looks like

Numbers in isolation do not tell a brand what to do first. Three points of sequencing follow from everything above.

Stock locally before optimising anything. Delivery speed feeds ranking on search led platforms, and no amount of listing optimisation compensates for a fulfilment time measured in weeks. This is an operational decision that has to precede the marketing one.

Build the review base where it is cheapest to build. If your category can be demonstrated, early volume through discovery led selling produces the social proof that search led ranking requires. Treating the two platforms as sequential phases rather than parallel channels tends to reach profitability sooner.

Establish verifiable presence off platform in parallel. This is the step most often deferred, and it is the one that decides whether the other two convert. A buyer who encounters a product in a video and cannot verify the brand anywhere else does not buy, and no amount of platform optimisation fixes an absence that sits outside the platform.

None of these depend on which platform is larger. They depend on where the brand currently stands, which is a different question and the one worth answering first.

The part that decides whether platform spend converts

There is a step between discovery and purchase that neither platform owns, and it is where foreign brands lose the most ground.

A buyer encounters your product in a video, becomes interested, and then leaves the platform to check whether the brand is real. They search the name. Increasingly they ask an AI assistant. If nothing credible comes back, the interest that the video paid for evaporates.

Arfadia's benchmark research puts a figure on both halves of this. Search accounts for 38.3 percent of brand discovery in Indonesia and social ads 37.3 percent, functionally tied. And six in ten Indonesians use social media as their primary channel for researching brands, which means the verification step happens across both.

Marketplace presence and search visibility are not competing budget lines. The second is what makes the first convert.

Frequently Asked Questions

Which is bigger in Indonesia, Shopee or TikTok Shop?

Shopee, on both measurements, but the gap depends on which you use. By access frequency Shopee holds 53.22 percent against TikTok Shop's 27.37 percent, per APJII and GoodStats August 2025. By 2024 GMV share, Shopee holds 46 percent against TikTok Shop's 11 percent, per Momentum Works via Databoks. The more useful question is which platform matches how your category is bought.

What is the Shopee and Tokopedia paradox?

Shopee is visited about 5.5 times more often than Tokopedia, 53.22 percent against 9.57 percent access frequency, but holds only roughly twice the GMV share, 46 percent against 23 percent. Tokopedia sessions carry substantially higher transactional value, which matters for higher AOV categories such as electronics, furniture and premium goods.

Do we need an Indonesian entity to sell on these platforms?

It depends on the platform and programme. Some marketplaces operate cross border seller programmes allowing listings without a local entity, but product categories requiring BPOM registration or halal certification still need an Indonesian entity or licensed local partner to hold those registrations.

How important is live commerce?

Video commerce reached 2.6 billion transactions in Indonesia in 2025, up 90 percent year on year, accounting for around 20 percent of online GMV, with roughly 800,000 sellers. Conversion runs about three times catalogue listings and 60 percent of buyers purchase during a live session. For consumer categories it is a primary discovery channel rather than an experiment, and it requires scheduling, hosts and inventory synchronisation.

Should we integrate GoPay, OVO and DANA separately?

No. QRIS, Bank Indonesia's unified QR standard, sits beneath those wallets and most bank applications, so one acceptance point reaches essentially the whole digitally active population. Integrating wallets individually is redundant work that produces a worse checkout. Foreign operators typically reach QRIS through licensed local aggregators.

What is the most expensive mistake?

Running Indonesia entirely from abroad. Local warehousing shortens delivery time, delivery time feeds marketplace ranking on search led platforms, and a product shipped from overseas on every order tends to lose visibility to a locally stocked competitor even at a better price.

Is search visibility still relevant if we sell inside marketplaces?

Yes, and it is often what decides whether marketplace spend converts. A buyer who encounters a product in content leaves the platform to verify the brand, by searching or by asking an AI assistant. Search accounts for 38.3 percent of Indonesian brand discovery and social ads 37.3 percent, and six in ten Indonesians use social as their primary brand research channel.

---