Carousell

1.1 Identity, origins, and scope

Carousell Group presents itself as a Singapore-founded, mobile-first marketplace built around the idea that selling should feel as easy as taking a photo and buying as easy as chatting. The current press fact sheet anchors the company in August 2012, describes a leading presence across Singapore, Hong Kong, Malaysia, Indonesia, the Philippines, Taiwan, and Vietnam, and frames the mission as making secondhand the first choice for everyone. The business has moved meaningfully beyond a single classifieds app: Carousell now describes a portfolio that includes regional classifieds brands, vertical assets in fashion, luxury, and mobile phones, and operating layers such as payments, shipping, seller subscriptions, and inspection-based trust services. The 2021 unicorn announcement still referenced eight markets and the OneKyat brand, while the 2026 fact sheet has shifted to seven markets and a different brand set. That suggests the company has already rationalised parts of its regional footprint rather than simply adding logos without pruning.[CO001, CO003, CO004, CO005, CO006, CO024]

1.2 Leadership, governance, and investor control points

The public record supports Siu Rui Quek as the central operating and narrative figure: he is repeatedly quoted as co-founder and CEO, he remains the spokesperson on funding, profitability, and IPO optionality, and the group's current transformation story is tied directly to his commentary. The named founding team in public databases includes Quek, Marcus Tan, and Lucas Ngoo, but disclosure on current executive depth is thinner than the strength of the brand might imply. The strongest newly visible non-founder leaders in 2026 are Shing Tai Leung, the chief strategy officer quoted on data and AI, and Rajath Ramesh, the product and platform engineering leader in Google Cloud's case study. JJ Ang appears as CFO in FY24 results coverage. Governance visibility remains incomplete: official materials confirm Jason Cho joining the board in 2021, while Tracxn lists additional board members, but committee structure, observer rights, and current investor designation are not public. That leaves Carousell with real key-person concentration risk around Quek even though the founder bench and investor roster are both broad.[CO002, CO023, CO028, CO030, CO032, CO033]

1.3 Economics, business-model shift, and scale indicators

The most important strategic fact about Carousell in 2026 is not simply that it remains a unicorn; it is that the company has materially changed the shape of the business since that 2021 mark. FY25 revenue reached US$141 million, up 18% year on year, and management said the group became EBITDA positive. Recommerce revenue grew 40% and accounted for 45% of total revenue, making it the biggest contributor to incremental growth. That stands in contrast to FY24, when public coverage described a slower topline year at US$119.3 million, a US$33.2 million net loss, and a cleaner but still incomplete view of the mix: US$53 million from recommerce, US$52.8 million from classifieds advertising, and US$13.5 million from other revenue. Public scale indicators still have caveats. Management uses rounded language such as tens of millions of users and 800+ teammates, while Dealroom shows 1,239 mapped employees. Those numbers are directionally helpful but not yet data-room grade.[CO007, CO008, CO009, CO014, CO015, CO016]

1.4 Milestones, adverse events, and current trajectory

Carousell's chronology shows a company that has alternated between capital formation, portfolio building, and operational tightening. The September 2021 US$100 million STIC-led round pushed valuation to US$1.1 billion and formally positioned the group around trusted secondhand commerce, including certified mobiles, auto programs, and integrated shipping. By 2025 the message had changed from growth-at-all-costs to disciplined execution: Quek emphasised capital optionality rather than imminent IPO timing, and industry coverage flagged both the aborted SPAC path and the December 2024 layoffs of 76 employees. At the same time the company deepened physical and AI-enabled recommerce. By end FY25 it operated 29 stores across four markets, claimed more than 20% GMV uplift in targeted categories from those stores, and described AI as already improving listing speed, fraud detection, translation, and engineering efficiency. The central diligence question is whether Carousell's FY25 EBITDA inflection reflects a durable new earnings profile or only the first step toward one.[CO010, CO011, CO012, CO013, CO019, CO020]

1.5 Exhibits

2.1 Market boundary and what counts

Carousell should not be underwritten against all Southeast Asian digital commerce. The right market boundary is the portion of consumer commerce where secondhand supply, peer-to-peer discovery, seller tools, and trust layers actually matter. Carousell's own materials show the company spanning goods, cars, luxury, mobile phones, fashion, property, home and living, and services; its recommerce strategy then adds inspection, seller subscriptions, integrated payments, shipping, and direct sell-to-platform workflows. That means the relevant market includes C2C and lightly intermediated secondhand transactions, classifieds monetization attached to those categories, and selected adjacent seller services. It excludes most first-party retail GMV, food delivery, pure grocery, and markets where Carousell does not operate. This distinction matters because broad SEA e-commerce TAM numbers are real but misleading if treated as directly addressable. Carousell is not competing for all online retail baskets; it is competing for the subset where trust, convenience, and local resale liquidity can beat both informal offline trade and horizontal mass-market platforms.[CM001, CM002, CM003, CM022, CM037]

2.2 Regional sizing: large backdrop, narrower serviceable market

Multiple third-party datasets support the conclusion that Southeast Asia is one of the world's fastest-growing digital commerce regions. UnivDatos sizes the 2024 market at US$201.9 billion, while AFFiNCO and Sellercraft place 2026 GMV around US$230-234 billion. Statista's topic page and Hashmeta's synthesis also point to a regional market already well above US$139 billion and still compounding at double-digit rates. The common picture is consistent even when the exact totals differ: Indonesia is the volume leader, Vietnam is among the fastest-growing markets, and Singapore is a smaller but far higher-spending premium market. For Carousell, this means the macro tide is supportive, but serviceable spend must be haircut aggressively for geography and category focus. Carousell does not cover Thailand, and much of headline regional GMV sits in categories or merchant models that look nothing like secondhand classifieds or recommerce. The right conclusion is not that Carousell owns a US$200+ billion market; it is that Carousell is riding a powerful macro shift inside a much narrower but still very large local resale and trust-enabled commerce wedge.[CM004, CM005, CM006, CM007, CM008, CM009]

2.3 Buyer segments, behaviors, and budget logic

The reviewed market sources consistently describe Southeast Asia as mobile-first, social-influenced, and review-driven. AFFiNCO estimates that 78% of e-commerce transactions now happen on mobile, while Sellercraft argues smartphone share is even higher. Hashmeta and AFFiNCO both emphasize review-reading, price comparison, and live-stream or influencer-led discovery as central habits, not edge behaviors. This environment fits Carousell unusually well. A secondhand marketplace needs users who browse on phones, compare listings across sellers, rely on reviews and badges, and accept wallet-based or on-platform payments. It also explains why Singapore, Indonesia, and Vietnam matter in different ways. Singapore is a premium market with higher order values and stronger expectations around convenience, while Indonesia and Vietnam offer greater volume and faster category growth. Budget ownership is mostly consumer household spend rather than corporate procurement, but the decision path varies by category: cars, luxury, and mobile devices require more trust and often more offline touchpoints than fashion or household goods. Carousell's value proposition therefore sits at the intersection of price sensitivity and risk reduction.[CM013, CM016, CM017, CM018, CM019, CM020]

2.4 Adoption drivers and what limits Carousell's share

The same regional trends that support Carousell also cap its upside. Mobile and wallet adoption, social discovery, and rapid e-commerce growth all help demand formation. Electronics, fashion, beauty, and other trust-sensitive categories also create room for differentiated recommerce, especially when inspection, warranties, and omnichannel handoffs reduce buyer fear. Yet the constraints are equally important. Shopee, Lazada, TikTok Shop, Facebook Marketplace, eBay, Mercari, and OLX all compete for parts of the same user job, either through scale, price, content-led discovery, or entrenched local liquidity. Offline retail still accounts for most spending, so Carousell must keep proving that its hybrid model solves trust better than ordinary listings do. And because many regional estimates are vendor-built rather than audited, it is easy to overstate market size or understate execution difficulty. Carousell's best path is not to mimic the largest marketplaces, but to dominate higher-friction secondhand use cases where trust infrastructure turns messy offline trade into repeatable digital behavior.[CM021, CM022, CM023, CM024, CM025, CM026]

2.5 Exhibits

3.1 Landscape by competitor class

Carousell should be benchmarked against a competitor set defined by user jobs, not just by other Southeast Asian startups. The closest substitute for many casual sellers is Facebook Marketplace, which combines discovery and built-in identity. For more structured resale and higher-scale seller behavior, eBay and Mercari are more relevant comparators because they put more emphasis on trust, tools, and repeat listing behavior. Traditional classifieds players such as OLX, Quikr, OfferUp, Craigslist, and Wallapop remain useful benchmarks for the persistence of local-listing demand, even if they do not match Carousell market for market. Depop and similar vertical specialists matter in fashion-led resale, while offline dealers and local social groups remain status-quo alternatives. This classification matters because Carousell is not trying to beat all rivals the same way. It needs a different answer for casual local resale, a different answer for higher-ticket trusted transactions, and another for vertical merchant workflows.[CP001, CP002, CP003, CP004, CP005, CP006]

3.2 Feature breadth and trust positioning

Carousell's strongest competitive argument is not raw assortment or social reach but trust-enabled completion. App-store surfaces and company materials show Buyer Protection, certified listings, delivery, and safer payment flows that go materially beyond bare-listing products. Singapore regulatory signals strengthen that posture further: Carousell publicly participates in enhanced verification measures and has a large visible safety-and-security help surface. Relative to Craigslist-like substitutes, that is a meaningful advantage. Relative to Facebook Marketplace, it is a more explicit trust stack but with less native discovery power. Relative to eBay, Carousell has a more regional and local-market focus but less global liquidity and less obviously mature seller tooling. The platform therefore sits in an intermediate position: stronger than lightweight classifieds on trust and completion, but not as broad as the largest global marketplaces or social networks.[CP008, CP009, CP010, CP011, CP014, CP022]

3.3 Pricing opacity, switching cost, and moat logic

Public competitor pricing is patchy, so capability and behavior evidence is more dependable than neat pricing charts. The most important structural fact is that multi-homing is easy. Sellers can often cross-list the same item across several apps, which keeps switching costs low and weakens any claim of hard lock-in. That makes liquidity density, trust signals, and repeat seller tooling more important than feature parity alone. Carousell benefits from verified flows, stores, and safer completion in trust-sensitive categories, but it is still vulnerable where buyers only need simple local listing and chat. The company's moat is therefore conditional. It is strongest where risk reduction changes user behavior enough to justify higher monetization or repeat use. It is weakest where listings are interchangeable, distribution is social-graph-driven, or price is the only important decision criterion. The practical implication for diligence is that management must prove category-level close-rate improvement, repeat seller behavior, and measurable trust-driven monetization rather than merely product breadth. Without those proofs, even a polished marketplace can still behave like a commodity listing surface.[CP012, CP013, CP015, CP017, CP018, CP019]

3.4 Adverse view: why competition can still compress outcomes

There is no evidence that Carousell enjoys winner-take-all control over recommerce in its markets. Larger ecosystems can imitate AI features, social platforms can intercept discovery before intent reaches Carousell, and free-listing substitutes can cap monetization in low-trust categories. Carousell's stores and verification stack are advantages, but they also add operating complexity that purely digital rivals do not bear. The investment question is therefore not whether competition exists; it clearly does. The question is whether Carousell can earn better trust-adjusted take rates, conversion, and repeat behavior in the categories it has chosen to emphasize. Current evidence supports a differentiated position, but not an unassailable moat. Investors should underwrite sustained rivalry and avoid assuming that product improvements alone eliminate marketplace commoditization. That adverse frame should shape valuation. Marketplace winners can still earn attractive returns without absolute dominance, but only if the chosen categories generate superior trust-adjusted economics. Carousell has evidence for that possibility, not proof that competition has been neutralized.[CP026, CP027, CP028, CP029, CP030, CP031]

3.5 Exhibits

4.1 Revenue model and mix

Carousell is no longer best described as a simple classifieds site monetized only by listing visibility or ad products. The strongest public evidence now shows a blended model: classifieds advertising, recommerce transactions, buyer-protected checkout, delivery-linked services, and merchant tools. FY24 reporting broke revenue into recommerce, classifieds advertising, and other revenue, while FY25 coverage emphasized that recommerce had grown to 45% of group revenue. App-store surfaces also reveal explicit monetization hooks such as CarouBiz subscriptions, listing bumps, and coins. This matters because it changes how revenue quality should be judged. Advertising-heavy classifieds businesses can be lighter but weaker on close-rate monetization, while transaction-led recommerce can compound take rate and repeat usage if trust, logistics, and inspection actually raise conversion. Carousell appears to be trying to move into that second category, but its public evidence still reveals more about topline direction than about gross-profit conversion by stream.[CI001, CI003, CI007, CI013, CI014, CI015]

4.2 Economics, costs, and operating leverage proxies

The clearest positive financial signal is that Carousell reached positive EBITDA in FY25 without abandoning growth altogether. Revenue rose from US$119.3 million in FY24 to US$141 million in FY25, while the company and independent coverage both framed recommerce as an increasingly important driver. Public sources also provide a few operational leverage hints. Google Cloud said AI-led listing tools reduced median listing time below two minutes and increased items listed by 20%, while automated translation now covers 90% of translation needs. The same case study says infrastructure migration cut compute-cluster cost by 10% and removed a large on-call burden. These facts do not equal full unit economics, but they do support the thesis that Carousell has found some real cost and productivity levers. The limitation is equally important: no public source in the reviewed set discloses take rates, contribution margins, CAC, or cohort profitability, so investors still cannot test whether EBITDA quality is durable across categories and markets. The operating story is better, but it is still not a substitute for direct category-level economics.[CI001, CI002, CI003, CI008, CI022, CI023]

4.3 Capital adequacy and financing dependency

Public evidence suggests Carousell is no longer in immediate capital-distress territory, but the current funding picture is still incomplete. Tracxn and The Company Check both show roughly US$376 million of cumulative equity financing, while multiple sources confirm the September 2021 US$100 million unicorn round at a US$1.1 billion valuation. Tracxn also records a smaller September 2023 extension led by GoTo. Management told The Straits Times in 2023 that the company was well capitalised and not reliant on an IPO as the only financing option, and FY24 cash of US$50.5 million offers one disclosed liquidity anchor. That said, the reviewed public package does not include FY25 ending cash, debt facilities, monthly burn, or store-level capital requirements. For underwriting purposes, this means investors can reasonably view funding dependency as improved relative to prior years, but cannot yet conclude that the business is fully self-funding or that external capital will be unnecessary under slower-growth or store-expansion scenarios.[CI006, CI009, CI010, CI011, CI012, CI032]

4.4 Financial verdict and diligence blockers

Carousell's financial arc is better than the stale “loss-making classifieds unicorn” narrative, but it still falls short of clean public-company-style transparency. The platform now has evidence of revenue scale, a more credible mix shift toward trust-led recommerce, and early operating leverage from AI and infrastructure work. Against that, almost every precision metric an investor would want for pricing or downside protection is still missing: take rates, gross margin by stream, repeat-purchase cohorts, seller CAC, cash conversion, and up-to-date runway. Public app pricing shows that monetization exists, but not how durable or profitable it is. The correct diligence stance is therefore constructive but conditional. Carousell appears to be improving the quality of its economics, yet the data package remains too sparse for a high-confidence intrinsic value call without direct management disclosure. That distinction should matter in committee discussions and price discipline.[CI021, CI024, CI029, CI030, CI031, CI033]

4.5 Exhibits

5.1 What the product delivers now

Carousell no longer presents only a listing board where buyers and sellers chat and leave the rest offline. The visible product stack now includes listing creation, reviews, safe checkout, official delivery, trusted home services, CarouBiz merchant tooling, Carousell Certified, and direct Sell to Carousell flows for selected categories. That broadening matters because it changes the job Carousell is solving. In lower-risk categories, the app can still behave like a lightweight peer marketplace. In higher-risk categories such as phones, luxury bags, and cars, the product increasingly inserts itself into verification, payment protection, and sometimes the inventory intake path itself. LuxLexicon and other category moves reinforce that Carousell is assembling multiple product lines around the same trust-heavy secondhand workflow rather than remaining a generic classifieds front end. That widening product scope explains why category operations and trust features now matter as much as listing liquidity alone.[CE001, CE002, CE003, CE004, CE005, CE006]

5.2 Architecture and operating model

Public partner evidence gives an unusually concrete picture of Carousell's technical stack for a private marketplace business. Google Cloud says the company uses Gemini, Vertex AI, BigQuery, Compute Engine, GKE, Cloud Storage, and Google Workspace across user-facing and internal workflows. The flagship “List with AI” feature uses multimodal input to generate title, description, category, price, and condition attributes, shifting listing from manual drafting to guided review. On the infrastructure side, Cloudflare describes a platform serving around one petabyte of images per month and handling flash-sale traffic spikes above three times normal volume with caching, WAF, DNS, and SSL services. This is not enough to reconstruct the whole architecture, but it is enough to conclude that Carousell's operating model is production-scale, cloud-native, and increasingly AI-mediated on both customer and internal workflows. The documented stack also implies meaningful vendor concentration, but it is still clearly beyond lightweight marketplace infrastructure.[CE011, CE012, CE013, CE014, CE015, CE016]

5.3 Differentiation, engineering signal, and roadmap

The most credible proof of product maturity is not just what Carousell says, but what its partners and public engineering surfaces show. GitHub reveals an organization with public repositories in Go, Swift, Kotlin, TypeScript, Python, and shell tooling, including the Orion microservice framework built around grpc/protobuf. Google Cloud also describes a roadmap that extends beyond the current AI listing experience toward sub-45-second listing time, richer stylistic attributes for discovery, and internal agentic tooling that may shorten engineering work. None of this proves a hard technology moat by itself. Larger competitors can buy or replicate many generic AI capabilities. But it does indicate a serious product and platform organization with the ability to turn infrastructure, data, and process improvements into user-facing speed and trust gains. Carousell's real edge appears operational and integrative rather than purely algorithmic. It also suggests product iteration is continuous rather than episodic, with infrastructure and ML changes feeding directly into the user workflow.[CE022, CE023, CE031, CE032, CE033, CE034]

5.4 Trust, safety, compliance, and quality controls

Trust and safety are central product features, not a compliance afterthought. Carousell's own help-centre materials describe AI and machine-learning detection, moderator review, dispute banners, real-time warnings, mobile verification, and 2FA or SMS re-verification. MHA and other independent reporting add stronger evidence still: Carousell committed to passkeys or biometric logins, account-suspension controls, and Singpass blacklist measures after verified accounts were abused for scams. The MHA transaction-safety page also frames user badges, secure payments, official delivery, and record retention as productized safeguards. Meanwhile, the sanctions policy shows a separate compliance layer dealing with unsupported geographies and prohibited transactions. Together these sources support a clear conclusion: Carousell has built a layered trust stack that spans interface, workflow, moderation, identity, and compliance. What remains unknown is the exact precision of those systems and the operational cost of running them at scale. In other words, Carousell is shipping a trust operating system as much as a marketplace app.[CE024, CE025, CE026, CE027, CE028, CE029]

5.5 Exhibits

6.1 Customer segments and who uses Carousell

Carousell's customer base is not one monolithic group. Public evidence supports at least four distinct user types: casual consumers clearing unused items, bargain-seeking buyers, specialist or professional merchants, and business sellers using CarouBiz. Category breadth matters here because each segment behaves differently. Home services, property, autos, and luxury are not just different products; they involve different trust needs, transaction speeds, and monetization potential. The existence of Specialist Awards, verified business badges, and CarouBiz terms all show that Carousell is consciously nurturing more structured seller types alongside everyday peer-to-peer use. At the same time, app-store language and the core product still position the platform as something ordinary consumers can use with minimal friction. That mix of casual and professional participation is central to how Carousell scales supply without becoming a pure merchant marketplace. In practice, each segment likely responds to a different mix of price, trust, convenience, and speed.[CU001, CU002, CU003, CU004, CU005, CU006]

6.2 Adoption trajectory and visible growth signals

The public package is much stronger on broad adoption than on exact growth mechanics. Carousell Group says it serves tens of millions of users across seven markets, while app-store surfaces speak of millions of people using the app to declutter, earn money, and source deals. The company does not currently publish a precise MAU figure in the reviewed sources, which limits trend precision, but it does provide several supporting adoption signals. These include 29 physical stores, strong category breadth, and an AI-led listing experience that Google Cloud says has already increased items listed by 20%. DealStreetAsia also reports more than 20% GMV uplift from stores and official trade-ins. That combination suggests the business is not just holding a large dormant user base; it is actively trying to increase transaction density, trust, and frequency through lower-friction listing and higher-trust offline touchpoints.[CU001, CU013, CU014, CU018, CU019, CU023]

6.3 Customer proof, trust signals, and satisfaction evidence

For a consumer marketplace, customer proof comes less from named enterprise logos and more from layered evidence about reviews, reputation signals, and actual transaction safeguards. Carousell's review help pages show a real reciprocal review system where buyers and sellers score each other on transaction quality and can leave compliments that become visible profile assets. Specialist Awards and verified business badges add another layer of merchant proof, especially for higher-intent buyers. The trust stack is reinforced by public scam and verification metrics: 99.96% scam-free transactions in 2025, more than 93% of users not encountering a scammer, and large-scale enforcement against suspicious accounts. The important caveat is that external customer sentiment is not perfectly clean. App-store ratings are strong, but review-aggregation sites still surface cautionary signals about safety and support quality. The right conclusion is that Carousell has meaningful proof of customer activity and trust, but not frictionless satisfaction. That distinction is important because marketplace health depends on many small trust cues compounding into willingness to transact again.[CU009, CU010, CU011, CU012, CU015, CU016]

6.4 Durability, expansion, and concentration risks

The durability question is where public evidence gets thinner. Carousell's likely land-and-expand path is not classic enterprise upsell but migrating users toward higher-trust and higher-monetization workflows: Buy button, certified categories, specialist merchants, Sell to Carousell, and stores. Those pathways can deepen engagement and raise revenue per active user, especially in phones, luxury, and autos. Customer concentration risk is also structurally different from enterprise software. Carousell is unlikely to depend on a handful of named accounts, but it can still be concentrated by geography, trust regulation, and seller mix. Singapore matters disproportionately because the most visible verification, Singpass, and anti-scam requirements are documented there. Most importantly, none of the reviewed public sources disclose actual retention, churn, cohort behavior, or professional-seller share. Investors therefore have a good scale story, a decent trust story, and an incomplete durability story. It also means management disclosure on cohorts would have outsized value relative to yet another broad user-count headline.[CU023, CU025, CU028, CU030, CU031, CU032]

6.5 Exhibits

7.1 Regulatory and legal risk

Regulatory pressure is not hypothetical for Carousell; it is already shaping product requirements in Singapore. MHA has directly assessed the platform under the Online Criminal Harms Act framework and required stronger controls after abuse of verified accounts persisted. That alone elevates regulatory risk above a generic consumer-platform concern. Legal and policy complexity is also broadening as Carousell handles sensitive identity, payment, location, chat, and dispute data across multiple jurisdictions. The sanctions and AML policy suggests a separate layer of compliance overhead around geography and financial-crime screening. None of this means the company is uniquely non-compliant, but it does mean risk is cumulative: scam prevention, identity verification, privacy, and cross-border compliance all converge in one customer-facing product. Because regulators can tighten requirements faster than platforms can cheaply adapt, compliance cost and user friction are both material downside channels. The terms and prohibited-content rules also show that Carousell must police behavior, listings, and identity at a granular level, increasing policy-enforcement burden as transaction volume scales.[CR001, CR002, CR003, CR004, CR005, CR006]

7.2 Operational, quality, and security risk

Carousell's operating risks are the classic problems of open marketplaces at meaningful scale: phishing, off-platform deal migration, deposit scams, counterfeit exposure, account takeover, and reliability expectations that rise as the platform pushes more transactions on-platform. The good news is that Carousell appears to run a layered mitigation stack, including ML detection, link blocking, review loops, official delivery, Buyer Protection, dispute banners, and moderation. The bad news is that every one of those protections costs money, adds workflow friction, or still fails in some edge cases. No public incident log or uptime disclosure resolves the availability question, and merchant terms explicitly carve out downtime beyond Carousell's control. The risk is therefore not simply that fraud exists; it is that the company may have to keep spending heavily on controls just to preserve trust at current levels, while still facing public criticism when scam counts rise.[CR009, CR016, CR017, CR018, CR019, CR020]

7.3 Dependency, financial, and market-structure risk

Carousell depends on a complex external stack. Google Cloud sits inside AI, data, and compute workflows, while Cloudflare supports core traffic handling and attack mitigation. App stores, payment systems, delivery partners, and verification rails also mediate parts of the customer journey. That would be manageable in a benign market, but the external market is anything but benign. Regional e-commerce and social-commerce are now dominated by a handful of scaled platforms, with Shopee and TikTok Shop exerting particularly strong pull on demand formation and seller economics. For Carousell, this means risk is both partner and market-structure driven: a cloud or platform problem can hurt operations, while an intensified competitive environment can force higher trust spend, lower take-rate ambition, or slower path to net profitability. EBITDA positivity is real, but it does not remove the model risk created by competitive intensity and still-incomplete public unit economics. In other words, Carousell is not competing in a calm classifieds niche anymore; it is underwriting trust and transaction quality inside a much harsher regional commerce stack.[CR021, CR022, CR023, CR025, CR026, CR028]

7.4 People, execution, and thesis-break triggers

Execution risk is still one of the most important investment risks because the strategy itself is operationally demanding. Carousell is simultaneously trying to increase trust, professionalize sellers, run stores and trade-ins, integrate acquisitions, and improve profitability. The company has already shown that miscalibration can lead to painful corrections, including workforce reductions in 2022 and 2024. This does not prove weak management; it does prove that the strategy has little room for sloppy execution. The founder-CEO remains the public face of capital allocation, profitability messaging, and strategic direction, which increases key-person sensitivity. Investors should therefore focus less on whether risk exists and more on whether it is being converted into monitorable controls. The thesis breaks if trust controls remain costly but ineffective, if category operations fail to earn their keep, or if competition forces Carousell back into subsidy-heavy behavior.[CR027, CR037, CR038, CR039, CR040, CR041]

7.5 Exhibits

8.1 Anchor valuation and business-quality shift

The starting point for Carousell valuation is still the September 2021 unicorn round at US$1.1 billion, because no cleaner public financing mark has replaced it. That anchor is old, but it is not irrelevant. In the years since, the company has moved from a still-loss-making but growing marketplace into a business that reported US$141 million of FY25 revenue and positive EBITDA. The operating profile is therefore better than the headline valuation timestamp implies. At the same time, investors cannot simply mark the company upward because business quality improved. The missing ingredients are public GMV, category-level take rates, durable net profitability, and evidence that recommerce stores and specialist verticals earn attractive returns. The right framing is not “old valuation equals wrong valuation,” but “old valuation now needs to be re-judged against stronger fundamentals and still-material uncertainty.” Investors also need to remember that private market marks can stay flat while fundamentals improve; that dynamic alone does not make the current price wrong, but it does make scenario analysis more important than headline nostalgia.[CV001, CV002, CV003, CV004, CV005, CV006]

8.2 Comparable set and multiple framing

No single comp is perfect. Vinted is the best premium private comp because it combines secondhand behavior, platform liquidity, integrated logistics and payments, and visible profitability. Back Market is useful as a higher-trust recommerce comp where service layers and trusted supply support richer economics. ThredUp is the best public transparency comp, even though public markets structurally discount it relative to private strategic winners. These peers suggest that Carousell should trade below the strongest private leaders but above a commodity or purely public-market mark if investors believe its transaction-led recommerce strategy is working. The challenge is that Carousell is simultaneously smaller than premium peers and more operationally ambitious than a thin marketplace. That combination argues for moderation, not maximalism, in valuation work. Geography matters too: Carousell competes in a faster-changing Southeast Asian commerce stack, whereas Vinted benefits from a broader European footprint and Back Market benefits from stronger refurbished-supply standardization.[CV009, CV010, CV011, CV012, CV013, CV014]

8.3 Valuation range and scenario analysis

A simple multiple framework is sufficient for a first-pass investment view. Public-style downside logic can justify a range around 3x to 4x revenue if the market insists on transparency, listed-comp discounts, and unresolved competition risk. A balanced private-market band of roughly 5x to 7x revenue reflects EBITDA positivity, category expansion, and an improving trust stack but still penalizes scale and disclosure gaps. A premium recommerce band of about 7x to 9x revenue becomes plausible only if investors view Carousell as a scarce regional winner with credible evidence that stores, specialists, and AI-enhanced trust tooling deepen moat rather than cost. The conclusion from this exercise is that US$1.1 billion is no longer obviously stretched, but it still sits near the optimistic end of a defendable 2026 range. In practice, this is a negotiation range rather than a single truth, and diligence quality should determine where inside the band an investor is willing to transact.[CV021, CV022, CV023, CV024, CV025, CV026]

8.4 Investment stance

The most defensible stance is a conditional-pass posture with an at-market to slightly above-market valuation view. Carousell has earned that better stance through real revenue scale, an EBITDA milestone, growing recommerce mix, and a strategy that now looks more structurally differentiated than it did in 2021. However, the company has not yet earned an unqualified premium mark close to the very best private recommerce leaders, because it still carries disclosure gaps, regional competition pressure, trust cost uncertainty, and execution complexity from stores and vertical operations. Investors should be open to the current valuation if management can prove durability rather than merely momentum. The next leg of upside depends less on narrative and more on whether FY26 converts EBITDA positivity into repeatable, risk-adjusted operating quality. Until those proofs arrive, discipline should come from terms, governance asks, and milestone-based conviction rather than from pretending that one point estimate is exact.[CV035, CV036, CV037, CV038, CV039, CV040]

8.5 Exhibits