Why Indonesian brands are sitting on the most complex O2O problem in the world
Indonesia is not just a large market — it is structurally one of the hardest online-to-offline measurement environments that exists. TikTok Shop, Shopee, and Tokopedia online. Indomaret, Alfamart, Guardian, and 17,000 islands of traditional trade offline. The gap between campaign signal and sell-through outcome is wider here than almost anywhere. That is precisely what makes it worth solving first.
Indonesia is not just a large market — it is structurally one of the hardest online-to-offline measurement environments that exists. TikTok Shop, Shopee, and Tokopedia online. Indomaret, Alfamart, Guardian, and 17,000 islands of traditional trade offline. The gap between campaign signal and sell-through outcome is wider here than almost anywhere. That is precisely what makes it worth solving first.
Most market entry arguments lead with size. Indonesia is the fourth most populous country in the world. The economy reached $1.4 trillion in GDP in 2024, ranking seventh globally by purchasing power parity. The retail market was valued at approximately $175 billion in 2024 and is projected to reach $278 billion by 2030. The e-commerce sector, valued at $75 billion in 2024, is growing at over 15% annually toward a projected $230 billion by 2032. Indonesia is now the third-largest e-commerce market globally, behind only China and the United States.
These numbers are real and they matter. But they are not why Indonesia is the right market for Veinera to enter first.
The reason is structural. Indonesia has, through a combination of geographic complexity, platform fragmentation, regulatory intervention, and channel multiplicity, created an online-to-offline measurement problem that is harder to navigate here than almost anywhere else in the world. And because the problem is harder, solving it is more valuable — and the advantage for brands that can do it is more durable.
The channel stack
To understand why Indonesia's O2O problem is uniquely complex, it helps to map what a consumer brand actually operates across when selling here.
Online: Shopee leads the Indonesian e-commerce market with approximately 52% share of regional GMV. The Tokopedia-TikTok Shop merged entity — created after TikTok's $1.5 billion investment in Tokopedia in December 2023 — is the second major platform, with TikTok Shop Indonesia reaching $6.2 billion in GMV in 2024, making it TikTok's second-largest market globally. Beyond Shopee and Tokopedia-TikTok, brands also operate across Lazada, Blibli, and Bukalapak, with the top three platforms collectively controlling over 85% of market share. Social commerce on TikTok and Instagram functions primarily as a discovery and demand-generation layer — purchase routing has been regulated separately since Ministerial Regulation 31/2023 banned direct in-app checkout on social media platforms, requiring purchase completion on separate marketplace platforms.
Modern trade offline: Convenience stores account for approximately 27% of organized retail sales in Indonesia. Indomaret and Alfamart dominate — Indomaret alone expanded over 1,000 new convenience outlets in 2025 across secondary cities. Guardian and Watson's serve the health and beauty segment in urban centers. Supermarkets and hypermarkets, though under pressure from e-commerce and convenience store competition, continue to serve a significant share of planned purchase occasions.
Traditional trade: This is the layer that makes Indonesia unlike most other markets. Traditional retailers — warungs, wet markets, independent grocery traders, and local distributors — still account for approximately 73% of Indonesia's grocery retail market, according to USDA data. They are the dominant channel for FMCG volume across second and third-tier cities, and across the 17,000 islands that make up the archipelago. Convenience store transactions exceeded 5 billion in 2024 in urban areas alone — but for the majority of Indonesian geography, the local warung remains the primary point of consumer purchase.
Payment and super-app ecosystem: Layered across all of these is a fragmented digital payments ecosystem — GoPay (71-88% usage rates), OVO (70-78%), DANA (61-83%), and ShopeePay (51-76%) — each embedded in its own super-app ecosystem spanning ride-hailing, food delivery, financial services, and commerce. These are not simply payment methods. They are distinct data environments with their own transaction records, loyalty signals, and behavioral profiles that do not automatically connect to one another.
A brand running a campaign in Indonesia is simultaneously managing signals across this entire stack. A TikTok campaign generates demand. Some of it routes to TikTok Shop via Tokopedia. Some routes to Shopee. Some converts in Indomaret. Some converts in a warung in Surabaya two weeks later. Some never converts online at all — it walks into a Guardian, looks at the shelf, and makes a decision.
Each of these outcomes is a legitimate commercial result of the campaign. Almost none of them can be traced back to the campaign that generated them through the data systems currently available to marketing teams.
What regulation did to the data problem
Indonesia's Ministerial Regulation 31/2023 — issued by the Ministry of Trade in September 2023 — banned direct transactional functionality on social media platforms. The regulation requires that platforms clearly separate social networking and purchasing capabilities. TikTok, which had been operating TikTok Shop with in-app checkout, was required to route purchases through a separate marketplace platform. The TikTok-Tokopedia merger was, in part, the structural solution to this regulatory constraint.
The regulation was designed to protect domestic small businesses from foreign platform dominance. Its commercial effect on the measurement environment was significant: it broke what might have been a more traceable demand-signal-to-purchase chain on social commerce, distributing the purchase completion step across a broader set of platforms and channels.
A consumer who discovers a product on TikTok and intends to buy it now has multiple completion routes: Tokopedia (the integrated marketplace), Shopee (the dominant platform), offline retail, or traditional trade. The discovery signal is on TikTok. The purchase completion is wherever the consumer happens to be most convenient at the moment of decision. The causal link between discovery and purchase is distributed across data environments that do not talk to each other.
This is not a problem unique to Indonesia — it is the O2O gap described throughout this series. But the regulatory intervention amplified it specifically in the social commerce context, at precisely the moment when social commerce was becoming the dominant demand-generation channel for consumer brands in the market.
The geography problem
Indonesia is 17,000 islands. This is a logistical reality that most market analyses acknowledge and then move past. For O2O measurement, it is a structural constraint that does not move past.
Geographic difference-in-differences — the causal inference method most suited to connecting regional campaign exposure to regional sell-through outcomes — relies on the ability to define treatment and control regions with meaningful geographic separation. In a continental market, this is relatively tractable. In an archipelago market where islands constitute natural geographic boundaries, the treatment-control structure has more natural clarity in some ways, but the data availability at the island or district level is correspondingly sparse.
Retailer sell-through data in Indonesia follows the concentration of modern trade. Data from Indomaret, Alfamart, and Guardian is available with reasonable coverage in urban and suburban areas, particularly across Java. Data from traditional trade — the 73% of grocery that moves through warungs and local distributors — is largely unavailable in standardized form. It is collected, if at all, through distributor sell-in reports, field sales audits, and periodic surveys — none of which are designed for campaign attribution.
This creates a measurement environment where the portion of the market where data is richest (modern trade, e-commerce) is also the portion that is smallest. And the portion where data is sparsest (traditional trade, second and third-tier cities, outer islands) is the portion that is largest.
Greater Jakarta held approximately 34% of Indonesia's retail market in 2025. But the cities and regions outside Jakarta are growing faster — Sulawesi alone is forecast to grow at an 8.75% annual rate through 2031. The market is moving toward the data-sparse regions. The measurement infrastructure is not.
The beauty and wellness case
Indonesia's specific product category dynamics make the O2O measurement problem particularly acute for the brands Veinera works with first.
Health and beauty commands approximately 14% of online retail and is the second-most-purchased category in Indonesian e-commerce after fashion, based on consumer survey data. On TikTok Shop, beauty and personal care is the dominant category — in Indonesia, 9 of the 10 top-grossing TikTok shops are from beauty and personal care. The Halal cosmetics and skincare segment has added structural demand as Indonesian consumers increasingly prioritize certified-halal products, aligning with the preferences of the country's predominantly Muslim population.
The brands that dominate this category — Wardah (ParagonCorp), Erha, Scarlett, and the established international brands that distribute through Indonesian modern trade — operate simultaneously on TikTok, Shopee, and Tokopedia online, and through Guardian, Watson's, Indomaret health corners, and pharmacy channels offline. A campaign for a Wardah skincare product that runs across TikTok creators and paid media generates demand that routes through all of these channels.
What the brand's analytics can see: TikTok performance metrics, Shopee sales, Tokopedia sales, paid media ROAS.
What it cannot see: Guardian sell-through in the week following the campaign. Indomaret reorder rates in the regions where the campaign's audience was concentrated. Distributor sell-in velocity to second-tier markets. Warung restocking patterns in areas where the brand's field team has relationships but no digital data trail.
The campaign ROI calculated from available data systematically undercounts the commercial outcome it produced. The decisions made from that ROI calculation are correspondingly distorted.
Why this complexity creates opportunity rather than an excuse to defer
The structural complexity of Indonesia's O2O environment is precisely the reason that solving it first is the right strategic choice, not a complication to be avoided.
Consider the alternative. An O2O measurement approach developed for a simpler market — one with consolidated retail channels, a single dominant e-commerce platform, and a compact geography — produces a methodology that works in that market and has limited generalizability. The assumptions built into the model reflect the market's simplicity.
An approach developed for Indonesia has to work across multiple competing e-commerce platforms, across modern trade and traditional trade simultaneously, across an archipelago geography with uneven data density, and in the presence of regulatory constraints that break the most obvious causal chains. A methodology that works in this environment works almost anywhere.
This is the argument for Indonesia-first that goes beyond market size. The analytical and commercial infrastructure built to close the O2O gap here is more defensible, more generalizable, and more capable of handling market complexity than infrastructure built for simpler conditions. Brands that develop this capability in Indonesia are better prepared for Europe, where GDPR constraints and fragmented offline retail create a different but structurally comparable measurement challenge, than brands that start in markets where the loop is easier to close.
Indonesia is not just a first market. It is a proving ground for a measurement approach that the global consumer brand industry needs but does not yet have.
What closing the loop requires here, specifically
Given the channel structure described above, closing the O2O measurement loop in Indonesia requires connecting four data environments that no existing system currently aggregates for campaign attribution purposes:
Creator and social campaign data — TikTok exposure, audience geography, creator performance metrics, and the content signals that drove engagement.
E-commerce platform data — Shopee, Tokopedia-TikTok, and secondary platform sell-through at the SKU and regional level, in time windows relevant to campaign attribution.
Modern trade sell-through data — Indomaret, Alfamart, Guardian reorder and sell-through data at the regional level, connected to geographic campaign exposure.
Distributor and traditional trade signals — Distributor sell-in velocity, field sales audit data, and regional demand indicators for the traditional trade channel that represents the majority of FMCG volume.
No single platform or analytics vendor connects all four. Each environment has its own data owner, its own cadence, its own format, and its own organizational home within the brand's structure.
The causal inference architecture that can connect them needs to work with imperfect data, at regional granularity, across an irregular geographic structure, with attribution windows that reflect the actual consumer decision timeline rather than the platform's default lookback period.
This is a technically demanding problem. It is also a commercially urgent one — and for the brands operating in Indonesia's consumer market, it is the difference between understanding what their campaigns actually produce and optimizing against the minority of their commercial outcome that their current systems happen to measure.
Sources and references
- MarkNtel Advisors. Indonesia Retail Market size: USD 175.11 billion in 2024, projected USD 278.29 billion by 2030.
- Technavio. Indonesia Retail Market Offline Segment: USD 154.4 billion in 2024.
- PS Market Research. Indonesian e-commerce market: USD 75.1 billion in 2024, 15.2% CAGR to USD 230.5 billion by 2032.
- US Trade.gov. Indonesia as third-largest e-commerce market globally.
- USDA. Retail Foods Annual, Indonesia 2025. Traditional retailers: approximately 73% of grocery retail market.
- Mordor Intelligence. Indonesia Retail Market report: Greater Jakarta 34.35% of Indonesia retail market 2025; Sulawesi forecast 8.75% CAGR through 2031; convenience stores 27% of organized retail.
- The Report Cubes. Indonesia Retail Market: Indomaret expansion of 1,000+ new outlets in 2025; convenience store transactions over 5 billion in 2024.
- TechBuzzChina / Momentum Works. TikTok-Tokopedia merger (December 2023); TikTok Shop Indonesia GMV USD 6.2 billion in 2024.
- Demeter ICT. Southeast Asia e-commerce: Shopee retaining 52% regional market share in 2024; Shopee-TikTok-Lazada collectively over 84% of regional market.
- Dwayne Gefferie / Substack analysis. Shopee-Tokopedia-TikTok-Lazada controlling 85%+ of Indonesian e-commerce market share; GoPay, OVO, DANA, ShopeePay usage rates.
- Indonesia trade.gov / Ministerial Regulation 31/2023. Regulation banning in-app transactional functionality on social media platforms, September 2023.
- Sellercraft. Indonesia e-commerce landscape and category analysis: health and beauty 14% of online retail, second-largest category.
- Momentum Works. Beauty and personal care dominance in TikTok Shop Indonesia.
