Carousell
FY25 EBITDA 拐点已现,但估值仍受信任与竞争风险约束
Carousell 的经营改善已足以让投资人认真推进接触:FY25 收入达到 US$141M, 循环电商已贡献 45% 收入,集团也报告 EBITDA 转正。 但公司仍夹在高溢价私有循环电商赢家与更便宜的上市二手交易可比公司之间; 信任、竞争和披露缺口也让 2021 年独角兽估值很难直接成立为高溢价案例。 合适姿态是有条件跟进:密切观察、加深尽调,只有 FY26 证明盈利经风险调整后仍可持续, 才上调信心。
封面要素
公司概况
Carousell 创立于新加坡,最初是面向消费者的移动优先市场平台,如今已从分类信息扩展为更宽的二手再商业平台,覆盖日常转售、垂直品类、认证商家、以旧换新和线下门店。公开证据显示,公司已具备实质经营规模——FY25 收入 US$141M,二手再商业占比 45%,EBITDA 转正——但它也仍在处理信任、披露和竞争复杂度,从高增长私营市场平台迈向更交易驱动的电商业务。
- 成立时间
- 2012-08-01
- 创始人
- Siu Rui Quek, Marcus Tan, Lucas Ngoo
- 创立地点
- Singapore
- 总部
- Singapore
- 产品
- Carousell 通过多品类市场平台变现,目前覆盖二手再商业交易、商品发布、专业卖家工具、垂直品类专家、以旧换新项目,以及支撑奢侈品、手机、时尚和其他高信任品类的线下门店层。
- 客户
- 覆盖新加坡、香港、马来西亚、印度尼西亚、菲律宾和台湾的消费者与重复卖家,以及使用可信交易流程的企业卖家和垂直商户。
- 商业模式
- 通过二手再商业交易、分类广告和商品发布产品、商户订阅 / 工具、垂直品类服务以及全渠道转售实现市场平台变现。
- 阶段
- Growth-stage private company
- 融资情况
- 最后一次清晰公开的定价事件,是 2021 年 9 月 Series E 独角兽轮,估值 US$1.1B;公开追踪平台显示累计融资约 US$376M。
执行摘要
主要优势
- FY25 收入 US$141M 且 EBITDA 转正,说明业务质量明显好过 2021 年独角兽标签本身传递的信息
- 循环电商现已贡献 45% 收入,模型更靠交易驱动,也可能比纯分类信息市场更守得住
- 29 家实体门店和专门品类流程,在奢侈品、时尚和手机品类里拼出差异化的全渠道信任主张
- 管理层明显在投入 AI 辅助信任、刊登和运营;如果也能降低摩擦和成本,护城河可能继续拉宽
- 相比眼前融资需求,公司资金看起来较充足,历史累计融资约 US$376M
主要风险
- 新加坡监管仍要求更强反诈骗控制,说明信任和验证风险还在塑造产品,而不是已经解决
- Shopee、TikTok Shop、Lazada 等大规模电商平台的竞争,可能挤压流量质量、卖家注意力和变现
- 公开披露仍缺 GMV、品类级抽佣率、FY25 净利润和门店级回报,限制了高估值论证的可信度
- 门店、专家、以旧换新和跨品牌整合会抬高运营复杂度;即便收入结构改善,执行负担也在加重
- 历史 US$1.1B 估值接近可辩护 2026 区间的乐观端,估值出错空间有限
未决问题
- 目前没有公开的 2026 年 Carousell 新股或老股定价事件,真实当前成交价仍未知
- 公开 GMV、贡献利润率、CAC 回本周期和门店队列经济性未披露
- FY25 期末现金、债务额度,以及欺诈或纠纷准备金假设未公开
- 如果 FY26 信任与安全支出或竞争强度进一步上升,EBITDA 能维持多久仍不清楚
- 专门垂直品类和实体门店的确切回报曲线,公开证据仍未证明
目录
01公司概览
1.1 身份、起点与业务边界
Carousell Group 把自己定位为创立于新加坡、移动优先的市场平台,核心理念是让卖东西像拍照一样简单,让买东西像聊天一样简单。当前新闻事实表把公司锚定在 2012 年 8 月,描述其在新加坡、香港、马来西亚、印度尼西亚、菲律宾、台湾和越南均有领先存在,并把使命表述为让二手成为每个人的第一选择。业务已明显不再只是一个分类信息应用:Carousell 现在描述的组合包括区域分类信息品牌、时尚、奢侈品和手机等垂直资产,以及支付、物流、卖家订阅、基于检测的信任服务等运营层。2021 年独角兽公告仍提到八个市场和 OneKyat 品牌,而 2026 年事实表已调整为七个市场和另一组品牌。也就是说,公司已经在梳理区域版图的一部分,而不是只加标识、不做取舍。[CO001, CO003, CO004, CO005, CO006, CO024]
公开 KPI 显示,Carousell 仍是私营公司,但运营更成熟;二手循环商业经济性强于单看过时 2021 年估值时的印象。
员工数和累计融资额不是管理层认证的当前 KPI;图中混合了官方声明、已报道业绩和数据库预览。
[CO007, CO008, CO009, CO010, CO019, CO022]1.2 领导层、治理与投资人控制点
公开记录支持把 Siu Rui Quek 视为核心运营和叙事人物:他多次以联合创始人兼 CEO 身份被引用,仍是融资、盈利和 IPO 可选性的发言人,集团当前转型故事也直接绑定他的评论。公开数据库列出的创始团队包括 Quek、Marcus Tan 和 Lucas Ngoo,但当前高管纵深的披露弱于品牌声量所暗示的水平。2026 年最强的新近可见非创始人领导者,是在数据和 AI 议题上被引用的首席战略官 Shing Tai Leung,以及 Google Cloud 案例研究中的产品和平台工程负责人 Rajath Ramesh。FY24 业绩报道中,JJ Ang 以 CFO 身份出现。治理可见度仍不完整:官方资料确认 Jason Cho 于 2021 年加入董事会,Tracxn 列出其他董事,但委员会结构、观察员权利和当前投资人席位安排并未公开。因此,即便创始团队和投资人名单都很宽,Carousell 围绕 Quek 仍存在真实的关键人物集中风险。[CO002, CO023, CO028, CO030, CO032, CO033]
| 人物 | 职务 | 背景 | 创始人-市场契合 / 职能覆盖 | 关键人依赖 |
|---|---|---|---|---|
| Siu Rui Quek | 联合创始人兼 CEO | 对外讲述融资、盈利、总部迁址和产品战略的人 | 从创立到独角兽轮、EBITDA 转正,保持直接连续性 | 关键——CEO 仍是资本、战略和运营姿态的主要发声人 |
| Marcus Tan | 联合创始人 | 公开数据库和早期公司资料将其列为创始团队成员 | 创始连续性,以及早期产品和社区搭建 | 中——当前运营职责公开披露不足 |
| Lucas Ngoo | 联合创始人 | 公司和数据库资料将其列为创始团队成员 | 创始连续性,以及早期产品和社区搭建 | 中——当前运营职责公开披露不足 |
| Shing Tai Leung | 首席战略官 | 在 FY25 业绩中就数据、AI 和信任优势发声 | 负责围绕十年买卖双方行为数据和 AI 部署搭建战略叙事 | 高——对当前 AI 驱动转型故事很重要,但并非唯一运营节点 |
| JJ Ang | CFO | 在 FY24 报道中就利润率纪律和更高利润率二手电商重点发声 | 亏损收窄期间负责财务控制和资本纪律 | 中——财务领导层可见,但完整财务团队未公开披露 |
该表只反映抓取材料中明确可见的管理者;完整董事会委员会和更广泛高管梯队仍部分未披露。
[CO002, CO020, CO023, CO030]| 利益相关方 | 角色 | 控制权或经济重要性 | 尽调请求 |
|---|---|---|---|
| Telenor Group | 大股东 / 战略支持方 | Dealroom 公开预览显示估算持股 26.2%,即便未披露为控制权仓位,也意味着影响力不小 | 请求最新股权结构表、股东协议,以及任何否决权或保留事项权利 |
| STIC Investments | 2021 年 Series E 领投方 | 领投 US$100M 独角兽轮,并为 Jason Cho 获得一个董事会席位 | 澄清当前持股、董事会权利,以及仍未清算的清算优先权或优先条款 |
| EDBI | 国资关联战略投资者 | 体现新加坡机构支持以及与可持续商业的契合 | 确认当前持股以及任何政策或战略合作预期 |
| Naver | 增长投资者 | 被列为主要投资者,并参与早期融资历史;该历史与 2020 年估值上调相关 | 请求 Naver 入股价格、后续跟投和任何商业合作的时间线 |
| Rakuten Ventures | 早期风险投资者 | 官方和数据库资料多次将其列为长期支持方 | 确认剩余持股,以及是否发生退出或二级转让 |
| Peak XV Partners | 原 Sequoia India 支持方 | 从高速增长阶段延续至今的风险投资支持方 | 澄清当前董事会或观察员身份,以及对最终流动性的意愿 |
| Quek / 创始人群体 | 管理和文化控制中心 | 叙事高度集中在 CEO 身上,意味着即便未披露投票权多数,也可能形成实际运营控制 | 请求创始人持股、归属安排和继任规划材料 |
经济重要性来自公开反复提及和 Dealroom 预览股权表推断,而非管理层提供的股东名册。
[CO021, CO022, CO023, CO028, CO029, CO030]1.3 经济性、商业模式转向与规模指标
2026 年关于 Carousell 最重要的战略事实,不只是它仍是独角兽,而是公司自 2021 年估值标记以来已实质改变业务形态。FY25 收入达到 US$141 million,同比增长 18%,管理层称集团已实现 EBITDA 转正。二手再商业收入增长 40%,占总收入 45%,成为增量增长的最大贡献者。这与 FY24 形成对比:公开报道当时描述的是收入增长更慢,收入 US$119.3 million、净亏损 US$33.2 million,收入结构更清楚但仍不完整:二手再商业 US$53 million,分类广告 US$52.8 million,其他收入 US$13.5 million。公开规模指标仍需打折。管理层使用数千万用户、800+ 团队成员这类取整表述,Dealroom 则显示 1,239 名映射员工。这些数字方向上有用,但还达不到尽调数据室级别。[CO007, CO008, CO009, CO014, CO015, CO016]
| 指标 | 数值 / 状态 | 日期 | 置信度 | 缺口 / 注意事项 |
|---|---|---|---|---|
| 成立 | 2012 年 8 月;新加坡 | 2012-08 | 高 | |
| 最新估值标记 | US$1.1B | 2021-09 | 高 | 未找到更新且佐证充分的公开估值标记 |
| 累计融资(数据库估算) | US$376M | 2026-06 | 中 | 数据库推算,非公司认证 |
| FY25 收入 | US$141M | FY2025 | 高 | |
| FY25 EBITDA | 为正 | FY2025 | 高 | EBITDA 绝对金额未公开披露 |
| FY25 二手电商占比 | 收入的 45% | FY2025 | 高 | |
| FY25 二手电商增长 | 同比增长 40% | FY2025 | 高 | |
| 线下门店 | 4 个市场 29 家 | FY2025 | 高 | |
| FY24 收入 | US$119.3M | FY2024 | 中 | 第三方基于业绩报道披露 |
| FY24 净亏损 | US$33.2M | FY2024 | 中 | 本轮未下载到直接申报文件 |
| FY24 现金 | US$50.5M | FY2024 | 中 | 来自行业报道,而非公司 IR 页面 |
| 用户 | 数千万 | 2026 | 高 | 公司使用取整表述;未披露精确 MAU |
| 团队规模 | 800+ 名团队成员 / 映射到 1,239 名员工 | 2026 | 低 | 公开来源中的当前员工数相互冲突 |
该表混合了当前官方表述、第三方 FY24 报道和数据库估算;FY25 精确现金、净利润和当前员工数仍是待尽调事项。
[CO001, CO007, CO008, CO009, CO010, CO014]1.4 里程碑、负面事件与当前轨迹
Carousell 的时间线显示,公司一直在资本形成、组合建设和运营收紧之间切换。2021 年 9 月,由 STIC 领投的 US$100 million 轮融资把估值推至 US$1.1 billion,也正式把集团定位在可信二手电商上,包括认证手机、汽车项目和一体化物流。到 2025 年,叙事已从不惜代价增长转为纪律化执行:Quek 强调资本可选性,而非迫近的 IPO 时间表;行业报道也提到终止的 SPAC 路径和 2024 年 12 月裁员 76 人。与此同时,公司加深了线下和 AI 驱动的二手再商业。到 FY25 年底,公司在四个市场运营 29 家门店,称目标品类的 GMV 因门店提升超过 20%,并表示 AI 已在提升发布速度、欺诈检测、翻译和工程效率。核心尽调问题是:Carousell 的 FY25 EBITDA 拐点,究竟反映出持久的新盈利画像,还是只迈出了第一步。[CO010, CO011, CO012, CO013, CO019, CO020]
| 日期 | 事件 | 类型 | 金额 / 估值 / 状态 | 参与方 | 含义 |
|---|---|---|---|---|---|
| 2012-08 | Carousell 在新加坡成立 | 创立 | 公司成立 | Siu Rui Quek、Marcus Tan、Lucas Ngoo 等创始人 | 确立新加坡起点和移动优先分类信息起步路径 |
| 2020 | Naver 领投增长融资,将估值推近 US$900M | 融资 | US$80M / ~US$900M | Naver 牵头的投资者财团 | 显示 2021 年估值跃升前,独角兽前风险已部分释放 |
| 2021-04 | Carousell Auto Group 和认证汽车业务成为重点 | 产品 | 项目启动 | Carousell Auto Group | 体现向更高信任度汽车交易垂直化 |
| 2021-09 | STIC 领投 Series E;达到独角兽里程碑 | 融资 | US$100M / US$1.1B | STIC;EDBI;现有投资者 | 提供最后一个佐证充分的估值标记,并刷新董事会 |
| 2021-09 | 融资新闻稿突出 Certified Mobiles 和 PosLaju 配送 | 产品 | 项目扩张 | Carousell;PosLaju | 显示便利性与信任论点从信息发布走向交易 |
| 2024-12 | 裁撤 76 名员工 | 不利 | 约 7% 裁员 | Carousell 管理层 | 证明公司控制成本,也显示需求与再投资之间的平衡并不均匀 |
| 2025-07 | FY24 业绩显示收入 US$119.3M,亏损收窄 | 规模 | US$119.3M 收入 / US$33.2M 净亏损 | Carousell 财务团队 | 证明公司在盈利里程碑前完成运营重置 |
| 2025-07 | Block 79 Ayer Rajah Crescent 新总部启用 | 治理 | 总部启用 | Carousell 领导层 | 即便保留 IPO 可选项,也释放更长期本地承诺 |
| 2026-07 | FY25 收入达 US$141M,EBITDA 转正 | 规模 | US$141M 收入 / EBITDA 为正 | Carousell Group | 是二手电商策略改善经济性的最强公开证据 |
| 2026-07 | 29 家线下门店和 AI 驱动的信任 / 发布工具成为叙事核心 | 产品 | 29 家门店 / 99% 无诈骗声明 | Carousell Group | 说明全渠道二手电商和 AI 信任栈已成为推介核心 |
2020 年融资里程碑使用二手媒体参考,纳入是为了保持估值连续性;2021 年轮次之外的确切月份和投资者财团构成仍是待尽调事项。
[CO001, CO010, CO019, CO021, CO022, CO032]Carousell 的公开轨迹从 2012 年新加坡创立出发,经过 2021 年独角兽融资,进入 2025-2026 年二手循环商业盈利拐点。
基于公开二级市场报道,2020 年融资里程碑被纳入叙事连续性;其他里程碑均在已采集来源中直接标注日期。
[CO001, CO014, CO019, CO021, CO022, CO032]Carousell 的公开叙事把创始人主导、分类广告流量、二手循环商业信任层和 AI 工具串成一个仍为私营公司的变现引擎。
[CO006, CO009, CO010, CO012, CO013, CO022]1.5 图表
02市场分析
2.1 市场边界与纳入范围
不能按整个东南亚数字商业来承销 Carousell。更合适的市场边界,是消费者电商中二手供给、点对点发现、卖家工具和信任层真正起作用的那一部分。Carousell 自身资料显示,公司覆盖商品、汽车、奢侈品、手机、时尚、房产、家居生活和服务;其二手再商业战略又加入检测、卖家订阅、一体化支付、物流,以及直接卖给平台的流程。这意味着相关市场包括 C2C 和轻中介二手交易、附着在这些品类上的分类信息变现,以及部分相邻卖家服务。它不包括大多数自营零售 GMV、外卖、纯生鲜杂货,以及 Carousell 没有运营的市场。这个区分很重要,因为宽口径东南亚电商 TAM 数字真实存在,但若视作可直接触达就会误导。Carousell 不是在争夺所有线上零售篮子;它争夺的是信任、便利和本地转售流动性能打赢非正式线下交易和横向大众平台的那一小块。[CM001, CM002, CM003, CM022, CM037]
| 视角 | 纳入支出 | 排除支出 | 为什么对 Carousell 重要 |
|---|---|---|---|
| Carousell 核心市场 | 二手商品、C2C 发布、本地转售、卖家服务、经验证的二手电商交易 | 自营零售、外卖、企业软件、纯物流 | 对应 Carousell 已有产品市场契合和变现抓手的领域 |
| 相邻市场 | 房产、汽车、服务和更高信任商业流程的分类信息 | 与交易无关的更广义数字广告 | 解释 Carousell 为什么不只在纯转售 SKU 中有价值 |
| 现状替代 | 线下跳蚤市场、经销商网络、本地 Facebook 群组、非正式转介绍 | 机构采购或批发贸易 | 说明 Carousell 既在和 App 竞争,也在和非结构化线下行为竞争 |
| 夸大 TAM 陷阱 | 东南亚电商总 GMV | Carousell 未服务的垂类或国家 | 避免把泛区域 GMV 数字直接当作 TAM |
边界是分析口径,不是公司口径:目标是把相关二手和分类信息支出,与更宽泛的区域电商背景分开。
[CM001, CM002, CM003, CM022, CM036, CM037]2.2 区域规模:大背景,较窄的可服务市场
多组第三方数据都支持一个结论:东南亚是全球增长最快的数字商业区域之一。UnivDatos 估算 2024 年市场规模为 US$201.9 billion,AFFiNCO 和 Sellercraft 则把 2026 年 GMV 放在约 US$230-234 billion。Statista 专题页和 Hashmeta 的综合也指向一个已远高于 US$139 billion、仍以双位数复合增长的区域市场。即便精确总量不同,共同图景仍一致:印度尼西亚是规模领头羊,越南位列增速最快市场,新加坡较小但客单和消费强度高得多。对 Carousell 而言,宏观潮水有利,但可服务支出必须因地理和品类聚焦大幅打折。Carousell 不覆盖泰国,许多头部区域 GMV 也落在与二手分类信息或二手再商业完全不同的品类或商户模式中。正确结论不是 Carousell 拥有一个 US$200+ billion 市场;而是 Carousell 乘上了一股强劲宏观迁移,但它所在的是更窄、仍然很大的本地转售和信任驱动电商楔形市场。[CM004, CM005, CM006, CM007, CM008, CM009]
| 视角 | 数值 / 区间 | 时间 | 定义 | 含义 |
|---|---|---|---|---|
| 广义东南亚电商 TAM | US$181B-US$234B | 2024-2026 | 第三方对区域电商总 GMV 或市场价值的估算 | 只适合作为外部上限;过宽,不能直接用于 Carousell 投资测算 |
| 2024 年东南亚电商市场 | US$201.9B | 2024 | UnivDatos 市场规模估算 | 锚定较保守的近期基线 |
| 2026 年东南亚电商 GMV | US$230B-US$234B | 2026 | Sellercraft 和 AFFiNCO 估算的 GMV 区间 | 支持宏观层面继续扩张 |
| Carousell 地理 SAM | 7 个运营市场的子集 | 当前 | 仅限 Carousell 当前已有规模运营的国家 | 将泰国和其他非核心市场剔出可触达支出 |
| Carousell 品类 SOM | 更高信任的二手品类 | 当前 | 奢侈品、手机、时尚、汽车及相邻本地转售场景 | 最相关的切入点,因为这些品类最受益于验货和信任层 |
该表有意混合广义区域估算和更窄分析切分,因为没有公开来源能干净隔离 Carousell 精确覆盖范围内的纯二手电商支出。
[CM004, CM005, CM006, CM008, CM009, CM013]Carousell 的相关市场从广义东南亚电商,收窄到七个市场中的二手与信任驱动商业楔子。
由于没有公开来源给出 Carousell 专属的二手循环商业 TAM,本口径用分析法逐层收窄。
[CM001, CM004, CM005, CM036]区域市场估算口径不同、水平不同,但都指向一个体量很大且仍在扩张的数字商业背景。
只有低端和高端端点直接来自公开发布;中点是根据来源估算选择的展示锚点。
[CM004, CM005, CM008, CM019, CM020]2.3 买家分层、行为与预算逻辑
所审阅的市场来源一致把东南亚描述为移动优先、受社交影响、由评价驱动的市场。AFFiNCO 估算,目前 78% 的电商交易发生在移动端,Sellercraft 认为智能手机占比更高。Hashmeta 和 AFFiNCO 都强调,看评价、比价格、直播或达人带动发现,是核心习惯而非边缘行为。这个环境格外适合 Carousell。二手市场平台需要用户在手机上浏览、跨卖家比较商品发布、依赖评价和徽章,并接受钱包或平台内支付。这也解释了新加坡、印度尼西亚和越南为什么以不同方式重要。新加坡是高端市场,订单价值更高,对便利性的要求更强;印度尼西亚和越南则提供更大规模和更快品类增长。预算所有权主要来自消费者家庭支出,而不是企业采购,但决策路径随品类变化:汽车、奢侈品和手机比时尚或家居用品更需要信任,也常需要更多线下触点。因此,Carousell 的价值主张落在价格敏感与风险降低的交汇处。[CM013, CM016, CM017, CM018, CM019, CM020]
| 客群 | 核心需求 | 用户行为 | 预算所有者 | Carousell 为什么能赢 |
|---|---|---|---|---|
| 价格敏感的日常买卖家 | 把闲置物品变现,也淘到便宜货 | 移动端优先浏览、比价、查评价 | 家庭消费预算 | 简单发布、本地聊天和市场流动性降低摩擦 |
| 高客单、重信任买家 | 在较低欺诈风险下购买二手手机、奢侈品或汽车 | 需要验真、配送和更安全的支付流程 | 决策更谨慎的家庭预算 | Certified、Buyer Protection、门店和卖给平台功能在这里更关键 |
| 小商家和专业卖家 | 经营重复发布或店铺 | 需要推广、分析和卖家工具 | 业主经营型微型企业预算 | CarouBiz 和卖家工具把场景从偶发 P2P 延伸出去 |
| 房产和汽车搜索者 | 重在线索生成和发现,而非即时结账 | 决策周期更长,会跨渠道交叉核验 | 家庭消费者预算或代理营销支出 | 即便最终交易转到线下,Carousell 作为分类信息入口仍有用 |
家庭消费支出占主导,但不同品类和信任要求下,决策路径差异明显。
[CM002, CM012, CM026, CM027, CM028, CM029]价格敏感度和信任需求都高的场景,最适合 Carousell。
[CM026, CM027, CM028, CM033, CM038]2.4 采用驱动因素与 Carousell 份额上限
支撑 Carousell 的区域趋势,也会限制其上行空间。移动和钱包普及、社交发现、电商快速增长,都有助于需求形成。电子产品、时尚、美妆和其他信任敏感品类,也为差异化二手再商业创造空间,尤其是检测、保修和全渠道交接能降低买家恐惧时。但约束同样重要。Shopee、Lazada、TikTok Shop、Facebook Marketplace、eBay、Mercari 和 OLX 都在争夺同一用户任务的一部分,手段可能是规模、价格、内容驱动发现或根深蒂固的本地流动性。线下零售仍占大部分支出,因此 Carousell 必须持续证明其混合模式比普通商品发布更能解决信任问题。而且许多区域估算来自供应商,而非经审计数据,很容易高估市场规模或低估执行难度。Carousell 最好的路径不是模仿最大市场平台,而是在高摩擦二手场景中占据主导;在那里,信任基础设施能把混乱的线下交易变成可重复的数字行为。[CM021, CM022, CM023, CM024, CM025, CM026]
| 因素 | 方向 | 类型 | 证据 | 含义 | 尽调请求 |
|---|---|---|---|---|---|
| 移动优先商业 | 正向 | 需求形成 | 审阅来源显示移动端占交易 78%+ 或更高 | 支持 App 原生发现和基于聊天的转售流程 | 请求按市场拆分 Carousell App 占 GMV 的比例 |
| 社交电商和直播发现 | 混合 | 发现 | TikTok Shop、Facebook Marketplace 和直播购物体量大且仍在增长 | 同时带来获客渠道机会和竞争压力 | 询问流量中有多少来自社交入口,相比直接访问或自然搜索占比如何 |
| 数字钱包与安全支付 | 正向 | 转化 | 钱包主导区域线上支付结构 | 支撑 Buyer Protection 与应用内结账渗透 | 索取 Buy button 在各品类、各市场的渗透率 |
| 线下信任仍重要 | 混合 | 品类行为 | 线下在 Sellercraft 中仍占零售支出的 80-85% | 支撑 Carousell 门店策略,但推高执行成本 | 追问各品类门店回本期和转化提升 |
| 高客单价二手品类 | 正向 | 品类结构 | 电子产品、时尚和高价值商品仍是区域线上重要品类 | 以信任为核心的二手交易可从非正式线下交易中抢份额 | 索取高客单价品类的 GMV 占比 |
| 横向平台竞争 | 负向 | 竞争 | Shopee、Lazada、TikTok Shop、Facebook Marketplace、eBay、Mercari、OLX 都覆盖部分用户需求 | 限制定价权,并推高 CAC 压力 | 追问按竞品、按品类拆分的胜负原因 |
| 过宽的 TAM 叙事 | 负向 | 方法论 | 区域来源采用不同定义和供应商估算 | 若不统一 TAM 口径,容易高估或低估 Carousell | 估值前先锁定一套投委会可用的 TAM 框架 |
本表有意混合需求因素和约束因素,因为 Carousell 的市场吸引力同时取决于 宏观增长和品类层面的信任摩擦。
[CM016, CM019, CM021, CM023, CM031, CM034]当区域消费者行为从发现推进到更安全地完成二手交易时,Carousell 受益。
[CM016, CM019, CM027, CM031, CM038]2.5 图表
03竞争对手
3.1 按竞争对手类型划分的格局
Carousell 的对标对象应由用户任务定义,而不只是其他东南亚创业公司。对许多随手卖东西的用户来说,最接近的替代品是 Facebook Marketplace,它把发现和内置身份结合在一起。对更结构化的转售和更大规模卖家行为而言,eBay 和 Mercari 更有参考意义,因为它们更强调信任、工具和重复发布行为。OLX、Quikr、OfferUp、Craigslist 和 Wallapop 等传统分类信息玩家,仍是判断本地商品发布需求韧性的有用基准,即便它们无法逐市场匹配 Carousell。Depop 和类似垂直专家在时尚驱动转售中重要,线下经销商和本地社群则仍是现状替代方案。这个分类很重要,因为 Carousell 不是要用同一种方式打败所有对手。面对随手本地转售,它需要一种答案;面对高客单可信交易,需要另一种答案;面对垂直商户工作流,还需要第三种答案。[CP001, CP002, CP003, CP004, CP005, CP006]
| 竞品 | 品类 | 规模 / 融资信号 | 目标客群 | 差异化 | 局限 |
|---|---|---|---|---|---|
| Facebook Marketplace | 社交 / 本地交易平台 | Meta 级分发 | 轻量本地买家和卖家 | 内置身份体系和发现信息流 | 已审公开页面上,交易信任栈不够显性 |
| eBay | 全球交易平台 | 上市公司级规模 | 结构化二手交易和广泛品类 | 全球流动性、买家信任和卖家研究工具 | 相比 Carousell,本地化和社区驱动较弱 |
| Mercari | 二手交易平台 | 主要二手平台 | 简单二手交易 | 强调“轻松转售” | 地理重心不同 |
| OLX / Quikr | 分类广告老牌平台 | 大型区域分类广告品牌 | 本地挂牌和发现 | 分类广告行为粘性强 | 公开信任 / 完成交易栈不够明显 |
| OfferUp / Craigslist / Wallapop | 本地挂牌替代品 | 根深蒂固的本地替代选择 | 快速本地转售 | 简单性和社区流动性 | 变现和安全深度不够清晰 |
| Depop | 垂直二手专家 | 聚焦时尚转售 | 风格驱动的年轻二手买家 | 品类专属社区和发现 | 品类范围更窄 |
画像比较的是买家要完成的任务,而不是严格的同市场重叠。
[CP001, CP002, CP003, CP004, CP005, CP006]3.2 功能宽度与信任定位
Carousell 最强的竞争论点不是原始供给量或社交触达,而是由信任驱动的成交。应用商店界面和公司资料显示,Buyer Protection、认证商品发布、配送和更安全支付流程,已明显超出裸商品发布产品。新加坡监管信号进一步强化了这个姿态:Carousell 公开参与强化验证措施,并有大面积可见的安全与保障帮助界面。相对 Craigslist 式替代品,这是有意义的优势。相对 Facebook Marketplace,它有更明确的信任栈,但原生发现力更弱。相对 eBay,Carousell 更聚焦区域和本地市场,但全球流动性更低,卖家工具成熟度也不如 eBay 明显。因此,平台处在中间位置:在信任和成交上强于轻量分类信息,但不如最大的全球市场平台或社交网络那样宽。[CP008, CP009, CP010, CP011, CP014, CP022]
| 购买标准 | Carousell | Facebook Marketplace | eBay | Mercari | 轻量分类广告平台 |
|---|---|---|---|---|---|
| 更安全支付或买家保护 | 已审应用页面显示支持 | 已审首页无法判断 | 整体支持,但本文未穷尽审阅 | 交易驱动模式支持 | 公开首页通常无法判断或能力有限 |
| 认证 / 验证流程 | 已审来源可见 | Unknown | 有品类专属信任项目,但不等同于 Carousell | 已审首页无法判断 | 通常较弱或不清晰 |
| 线下或辅助式信任触点 | 门店和卖给平台流程 | 已审页面无证据 | 主要是数字化 | 主要是数字化 | 主要是数字化 |
| 卖家工具深度 | 中等 | Unknown | 高,含研究工具 | 中等 | 低到未知 |
| 社交发现触达 | 低于 Meta | 很高 | 中等 | 中等 | 低到中等 |
证据不足的单元格标为未知,而不是靠猜。
[CP008, CP011, CP013, CP014, CP022, CP023]在信任和成交深度上,Carousell 介于轻量分类广告与更大的交易驱动型平台之间。
坐标轴是证据支持的序数评分,不是审计过的数字基准。
[CP002, CP003, CP004, CP008, CP011, CP019]能力对比上,Carousell 优于纯刊登,但不必然胜过最大的横向交易平台。
未知单元格按保守口径处理;矩阵使用已审阅公开材料中的序数证据。
[CP008, CP013, CP014, CP022, CP023, CP018]3.3 定价不透明、切换成本与护城河逻辑
公开竞争对手定价资料零散,因此能力和行为证据比整齐的价格表更可靠。最重要的结构性事实是多平台发布很容易。卖家常能把同一件商品跨多个应用发布,这让切换成本保持低位,也削弱任何硬锁定主张。于是,流动性密度、信任信号和重复卖家工具比单纯功能对齐更重要。Carousell 在信任敏感品类中受益于认证流程、门店和更安全成交,但在买家只需要简单本地商品发布和聊天的场景里仍然脆弱。因此,公司的护城河是有条件的。风险降低足以改变用户行为、支撑更高变现或重复使用时,它最强;商品发布可替代、分发由社交图谱驱动、价格是唯一重要决策标准时,它最弱。对尽调而言,实际含义是管理层必须证明品类层面的成交率改善、重复卖家行为和可衡量的信任驱动变现,而不能只证明产品宽度。没有这些证明,即便市场平台打磨得很漂亮,也仍可能像商品化商品发布界面一样运转。[CP012, CP013, CP015, CP017, CP018, CP019]
| 平台 | 公开定价可见度 | 观察到的模式 | 未知项 | 启示 |
|---|---|---|---|---|
| Carousell | 部分 | 挂牌加信任和卖家服务追加销售 | 实际抽佣率和订阅兑现因市场而异 | 公开界面显示存在多元变现 |
| Facebook Marketplace | 低 | 可能主要靠生态分发变现,而非显性的挂牌费 | 已审页面无法看清本地卖家费用或推广机制 | 难以用简单交易平台费率表对比 |
| eBay | 部分 | 交易驱动型平台,配套卖家工具 | 实际抽佣率因品类和卖家而异 | 卖家工具成熟度或能支撑更好变现 |
| Mercari | 低到部分 | 交易驱动的二手模式 | 此处未审阅品类和市场定价细节 | 比免费分类广告更接近的战略参照 |
| 分类广告同行 | 低 | 许多市场可能靠置顶、广告或订阅 | 公开定价通常要开户后才可见 | 能力对比比价格对比更稳健 |
定价不透明本身就是竞争事实,因为它限制了公开层面干净的变现比较。
[CP020, CP021, CP024, CP025]Carousell 的信任护城河可信,但转换成本低,护城河能否兑现仍有前提。
KPI 数值是基于来源证据的定性评分,不是公司披露指标。
[CP017, CP018, CP019, CP028, CP030, CP032]3.4 反向视角:竞争为何仍会压缩结果
没有证据显示 Carousell 在其市场中的二手再商业拥有赢家通吃控制力。更大生态可以模仿 AI 功能,社交平台可以在意图到达 Carousell 前截走发现,免费商品发布替代品可以限制低信任品类的变现。Carousell 的门店和验证栈是优势,但也带来纯数字竞争对手无需承担的运营复杂度。因此,投资问题不是竞争是否存在;竞争显然存在。问题是 Carousell 能否在自己选择强调的品类中,获得更好的信任调整后抽佣率、转化和重复行为。当前证据支持差异化定位,但还不支持不可攻破的护城河。投资人应承销持续竞争,避免假设仅靠产品改善就能消除市场平台商品化。这个反向框架应塑造估值。市场平台胜者即便没有绝对统治,也可以赚取有吸引力的回报,但前提是所选品类产生更优的信任调整后经济性。Carousell 有这种可能性的证据,但没有竞争已被中和的证明。[CP026, CP027, CP028, CP029, CP030, CP031]
| 护城河主张 | 威胁 | 严重性 | 缓释 / 证据 | 尽调追问 |
|---|---|---|---|---|
| 高风险品类中的可信交易 | 更大平台复制数字信任功能 | 高 | 门店、认证和监管验证可以加深差异化 | 索取可信流程带来的品类级转化提升 |
| 本地市场流动性 | 多平台并用稀释独占性 | 高 | 若成交率提升,品牌和信任仍能发挥作用 | 索取重复卖家率和独家挂牌占比 |
| AI 辅助的平台质量 | 大平台功能追平 | 中 | AI 可能改善运营,但不太可能单独构成护城河 | 索取可归因于 AI 的滥用减少和转化提升 |
| 监管信任姿态 | 规则可能扩散到所有同行 | 中 | 早合规仍能建立品牌信任 | 追问验证成本按市场如何变化 |
| 线下门店网络 | 执行复杂度和成本 | 中 | 在手机、奢侈品等重查验品类中可能重要 | 追问门店回本期和品类提升 |
风险登记表衡量耐久度,而不只是差异化。
[CP015, CP017, CP019, CP026, CP027, CP028]3.5 图表
04财务
4.1 收入模式与结构
Carousell 已不再适合被描述为一个只靠商品发布可见度或广告产品变现的简单分类信息网站。最强公开证据显示,它现在是混合模式:分类广告、二手再商业交易、买家保护结账、与配送挂钩的服务,以及商户工具。FY24 报告把收入拆为二手再商业、分类广告和其他收入;FY25 报道则强调二手再商业已增长至集团收入的 45%。应用商店界面也显示了明确变现入口,如 CarouBiz 订阅、商品曝光加速和 coins。这很重要,因为它改变了判断收入质量的方式。广告重的分类信息业务可以更轻,但在成交率变现上更弱;交易驱动的二手再商业若信任、物流和检测确实提升转化,就能复合抽佣率和重复使用。Carousell 似乎正试图进入第二类,但其公开证据对收入方向的揭示,仍多于对各收入流毛利转化的揭示。[CI001, CI003, CI007, CI013, CI014, CI015]
| 收入流 | 机制 | 单位 | 当前数值 / 状态 | 质量判断 | 尽调追问 |
|---|---|---|---|---|---|
| 二手交易 | 直接交易、查验、回购和信任驱动服务 | 收入 | FY25 收入的 45%;同比增长近 40% | 质量在改善,因为收入绑定交易完成,而不只是流量 | 索取按收入流拆分的品类和地域组合及毛利率 |
| 分类广告 | 挂牌推广和类似广告的商户支出 | 收入 | FY24 为 US$52.8M | 有用,但确定性可能低于与交易挂钩的变现 | 索取按商户层级拆分的附加率和流失率 |
| 其他收入 | 未说明的附属收入流 | 收入 | FY24 为 US$13.5M | 体量小,但说明已超出两个头部收入桶的多元化 | 索取确切构成和利润率结构 |
| 卖家订阅和应用内工具 | CarouBiz、Coins、Bumps | 标价 / 订阅 | 公开定价可见;实际收入未知 | 支撑变现广度,但不能证明实际兑现收益 | 索取付费商户数和 ARPU |
| 有买家保护的结账和配送 | 类托管支付和履约支持 | 交易服务 | 应用商店描述中公开可见;收入未披露 | 若提升成交率和信任,质量可能更高 | 索取 Buy button 渗透率和配送抽佣率 |
公开证据最强的是收入方向,最弱的是每单位活动的实际经济性。
[CI003, CI007, CI013, CI014, CI018, CI019]| 项目 | 公开价格 / 模式 | 标价与实际兑现 | 未知项 | 来源 | 启示 |
|---|---|---|---|---|---|
| CarouBiz Lite | US$14.99 | 仅标价 | 采用率、折扣和特定市场定价未知 | Apple App Store | 确认 SMB 卖家订阅变现存在 |
| Carousell Coins | US$2.99 至 US$12.99 | 仅标价 | 用途和收入分成未知 | Apple App Store | 显示微交易变现 |
| 3-Day Bump | US$2.99 | 仅标价 | 频次和转化提升未知 | Apple App Store | 可见的付费推广机制 |
| Buyer Protection / Buy button(买家保护 / 购买按钮) | 可能是交易挂钩变现 | 模式可见,经济性不可见 | 抽佣率、退款和保护成本未知 | Google Play | 暗示更高质量的变现路径 |
| Official Delivery | 可能是服务挂钩变现 | 服务可见,经济性不可见 | 合作方经济性和损失率未知 | Google Play | 增加履约相关收入机会 |
公开定价界面是局部的,且很可能没有覆盖所有市场。
[CI014, CI015, CI016, CI017, CI018, CI019]Carousell 现在不只靠列表曝光变现,而是在转售交易的多个环节收费。
[CI013, CI014, CI018, CI019, CI020]4.2 经济性、成本与经营杠杆代理指标
最清晰的正面财务信号是,Carousell 在 FY25 实现 EBITDA 转正,同时没有完全放弃增长。收入从 FY24 的 US$119.3 million 升至 FY25 的 US$141 million,公司和独立报道都把二手再商业描述为越来越重要的驱动因素。公开来源还提供了一些运营杠杆线索。Google Cloud 称,AI 驱动的商品发布工具把中位发布时间降至两分钟以内,并让上架商品数增加 20%;自动翻译现在覆盖 90% 的翻译需求。同一案例研究称,基础设施迁移把计算集群成本削减 10%,并移除了大量值班负担。这些事实不等同于完整单位经济,但支持一个判断:Carousell 找到了一些真实的成本和生产率杠杆。限制同样重要:所审阅公开来源没有披露抽佣率、贡献利润率、CAC 或分群盈利能力,因此投资人仍无法测试 EBITDA 质量是否能跨品类和市场持久。经营故事变好了,但仍不能替代直接的品类层面经济性。[CI001, CI002, CI003, CI008, CI022, CI023]
| 指标 | 数值 / null | 置信度 | 重要性 | 尽调追问 |
|---|---|---|---|---|
| 按收入流拆分的毛利率 | Null | 低 | 二手交易相对分类广告更好或更差,取决于查验、退货和物流成本 | 索取分类广告、二手交易和其他收入的毛利率拆分 |
| 按品类拆分的抽佣率 | Null | 低 | 显示变现能力和信任溢价 | 索取手机、奢侈品、时尚、汽车和服务的抽佣率 |
| 卖家 CAC | Null | 低 | 用于检验获客效率能否对上网络效应 | 索取按市场拆分的付费和自然卖家获取 |
| 回本期 | Null | 低 | 决定增长能否自我供血 | 索取按商户层级和品类拆分的队列回本期 |
| 可信交易贡献利润率 | Null | 低 | 检验门店和查验是否带来经济增量 | 索取每个认证品类的贡献利润率 |
| EBITDA 状态 | FY25 转正 | 中 | 显示成本纪律和规模出现台阶式改善 | 索取 EBITDA 金额及通往经营现金流的桥接 |
本表有意保留大量缺口,因为本章应保留公开证据无法支撑的内容。
[CI002, CI021, CI022, CI023, CI024, CI037]公开证据支持经营杠杆改善,但核心单位经济桥仍未闭合。
由于公开数据未披露毛利率或抽佣率,这条链路只能作方向性判断。
[CI022, CI023, CI026, CI037]4.3 资本充足性与融资依赖
公开证据显示,Carousell 已不再处于即时资本困境,但当前融资图景仍不完整。Tracxn 和 The Company Check 都显示累计股权融资约 US$376 million,多个来源确认 2021 年 9 月 US$100 million 独角兽轮,估值 US$1.1 billion。Tracxn 还记录了 2023 年 9 月由 GoTo 领投的一笔较小延展轮。管理层在 2023 年告诉 The Straits Times,公司资本充足,不依赖 IPO 作为唯一融资选项;FY24 现金 US$50.5 million 则提供了一个已披露的流动性锚点。话虽如此,所审阅公开资料包不包括 FY25 期末现金、债务额度、月度烧钱,或门店层面资本需求。承销时,投资人可以合理认为融资依赖相对前几年已有改善,但还不能得出业务已完全自我融资,或在低增长、门店扩张情景下不需要外部资本的结论。[CI006, CI009, CI010, CI011, CI012, CI032]
| 项目 | 公开数值 / 状态 | 日期 | 置信度 | 重要性 | 尽调追问 |
|---|---|---|---|---|---|
| 现金及现金等价物 | US$50.5M | FY24 | 中 | 已审公开来源中最新披露的流动性锚点 | 索取 FY25 期末现金和本月现金桥接 |
| 总股权融资 | ~US$376M | 截至 2023 年 9 月累计 | 高 | 体现长期投资人支持,也说明有能力为转向继续输血 | 索取股权结构表和清算优先权结构 |
| 最近可见融资 | GoTo 领投的 US$2.8M Series E 延伸轮 | Sep 2023 | 中 | 说明成为独角兽后仍有支持,但规模更小 | 确认该轮偏战略、过桥,还是员工套现 |
| 最后公开估值 | US$1.1B | Sep 2021 | 高 | 估值纪律的关键参照点 | 索取 2021 年以来任何内部或外部估值标记 |
| 资金续航 / 债务义务 | 无 / 未披露 | 当前 | 低 | 偿付能力和下行风险的核心变量仍缺失 | 索取债务、担保、库存融资和资金续航情景 |
资本充足度看起来已有改善,但仅靠公开数据仍无法完全核验。
[CI006, CI009, CI010, CI011, CI012, CI032]公开财务图景显示轨迹改善,但许多 2026 年资产负债表数值仍未披露。
盈利能力的中位和高位值只是方向性估计,因为 FY25 EBITDA 金额未公开。
[CI001, CI002, CI005, CI006, CI033, CI034]4.4 财务结论与尽调阻断项
Carousell 的财务弧线已经好过“亏损分类信息独角兽”这个过时叙事,但仍达不到上市公司式透明度。平台现在有收入规模证据,更可信地转向信任驱动的二手再商业,并从 AI 和基础设施工作中初步释放经营杠杆。与之相对,投资人用于定价或下行保护几乎需要的每个精确指标都仍然缺失:抽佣率、按收入流划分的毛利率、重复购买分群、卖家 CAC、现金转化,以及最新现金续航。公开应用定价显示变现存在,但没有显示它有多持久、多赚钱。因此,正确的尽调姿态是建设性但有条件。Carousell 的经济质量似乎在改善,但如果没有管理层直接披露,数据包仍过于稀疏,无法给出高置信度内在价值判断。这个差异应进入投委会讨论和价格纪律。[CI021, CI024, CI029, CI030, CI031, CI033]
| 缺失指标 | 影响 | 仍然缺失的原因 | 具体尽调路径 |
|---|---|---|---|
| FY25 期末现金 | 无法检验资金续航 | 对外公告省略资产负债表细节 | 索取最新月度管理报表 |
| EBITDA 金额和利润率 | 盈利质量不清楚 | 公开资料称 EBITDA 为正,但未说明金额 | 索取 EBITDA 桥表,以及调整后与未调整口径定义 |
| 按收入流拆分的毛利率 | 估值质量受损 | 公开资料未披露收入流层面的毛利率 | 索取经审计分部或内部管理口径 |
| 抽佣率和交易渗透率 | 无法判断信任驱动变现的质量 | 公开定价只显示工具,不显示实际经济性 | 索取各市场购买按钮渗透率和抽佣率 |
| 卖家获客和留存 | 无法判断 CAC / 回本周期 | 未披露队列或漏斗数据 | 索取商家漏斗和重复卖家队列 |
| 门店层面经济性 | 运营强度不确定 | 实体门店数量已公开,回本周期未公开 | 索取各品类门店贡献毛利和回本周期 |
这些缺口说明,证据更能支持方向判断,而不是精确测算。
[CI021, CI022, CI033, CI036, CI037, CI038]二手电商扩大了变现范围,但也比传统分类信息引入更高的运营强度。
矩阵采用分析口径,不是公司披露,用来显示更高质量的收入可能需要更重的运营。
[CI020, CI023, CI024, CI029, CI038]4.5 图表
05产品与技术
5.1 当前产品交付内容
Carousell 已不再只呈现一个买卖双方聊天、其余留在线下完成的商品发布看板。可见产品栈现在包括商品发布创建、评价、安全结账、官方配送、可信家政服务、CarouBiz 商户工具、Carousell Certified,以及部分品类的直接 Sell to Carousell 流程。这个拓宽很重要,因为它改变了 Carousell 正在解决的任务。在较低风险品类中,应用仍可像轻量点对点市场平台一样运转。在手机、奢侈包和汽车等较高风险品类中,产品越来越多地插入验证、支付保护,有时还插入库存收货路径本身。LuxLexicon 和其他品类动作进一步说明,Carousell 正围绕同一套信任重的二手流程拼装多条产品线,而不是停留在通用分类信息前端。产品范围扩大,也解释了为什么品类运营和信任功能如今与商品发布流动性同样重要。[CE001, CE002, CE003, CE004, CE005, CE006]
| 模块 / 资产 | 主要用户 | 状态 / 成熟度 | 差异化 | 尽调缺口 |
|---|---|---|---|---|
| Marketplace 上架和聊天 | 所有买家和卖家 | 成熟 | 快速、本地、移动优先的上架和议价核心 | 需要按品类拆分活跃用户和转化 |
| CarouBiz 卖家工具 | 高频卖家 / 商家 | 成熟但不透明 | 覆盖 app 和网页的订阅制商家功能层 | 需要按市场和套餐拆分的完整功能表 |
| Buyer Protection / 购买按钮 | 买家和符合条件的卖家 | 在部分市场成熟 | 平台内安全结账和纠纷处理 | 需要渗透率、退款率和抽佣率 |
| Carousell Certified | 高信任品类买家 | 在部分品类成熟 | 检测为手机、奢侈品和汽车交易建立信心 | 需要缺陷率和成交率提升 |
| Sell to Carousell(卖给 Carousell) | 追求速度或确定性的卖家 | 在部分品类增长 | 面向服装、手机、奢侈包和汽车的直接收货流程 | 需要利润率和周转效率 |
| LuxLexicon / 垂直品牌 | 奢侈品品类用户 | 增长中 / 品类特定 | 奢侈品转售里的多品牌专精 | 需要与主 app 的整合深度和重叠度 |
矩阵聚焦可见产品模块,而不是后台团队。
[CE001, CE003, CE004, CE005, CE006, CE008]| 用户任务 | 当前工作流 | Carousell 方案 | 可衡量收益 | 局限 |
|---|---|---|---|---|
| 快速发布一件商品 | 拍照、写文案、手动定价 | AI 辅助发布会预填标题、描述、品类、价格和成色 | 上架时间中位数低于两分钟 | 需要品类级准确率统计 |
| 购买风险更高的二手商品 | 找商品、比卖家,还要担心真伪或成色 | 认证商品、评价、Buyer Protection、官方配送 | 信任更高,纠纷路径更清晰 | 具体转化提升未知 |
| 即时卖出或托管转售 | 手动和买家议价 | 部分品类提供 Sell to Carousell / 寄售式流程 | 回款更快,议价负担更轻 | 经济性和周转时间未公开 |
| 经营小型转售生意 | 反复发布并推广库存 | CarouBiz 订阅、金币、置顶、卖家功能 | 超出轻度 P2P 使用的专用商家工具 | 各套餐功能深度不透明 |
| 在平台上安全交易 | 避开钓鱼和假货风险 | 链接和二维码拦截、警示、徽章、验证、审核 | 诈骗暴露更低,决策支持更好 | 误报和客服成本未公开 |
收益来自公开资料或方向性判断;经济结果仍需要管理层数据。
[CE002, CE009, CE010, CE012, CE013, CE025]产品把转售流程从上架压缩到可信履约。
[CE002, CE004, CE005, CE009, CE029]5.2 架构与运营模式
对一家私营市场平台而言,公开伙伴证据给出了少见具体的 Carousell 技术栈图景。Google Cloud 称,公司在面向用户和内部工作流中使用 Gemini、Vertex AI、BigQuery、Compute Engine、GKE、Cloud Storage 和 Google Workspace。旗舰 “List with AI” 功能使用多模态输入生成标题、描述、品类、价格和成色属性,把商品发布从手工撰写转为引导式审核。在基础设施侧,Cloudflare 描述的平台每月提供约 1 PB 图片,并用缓存、WAF、DNS 和 SSL 服务处理超过正常量三倍的闪购流量峰值。这些信息不足以重建完整架构,但足以得出结论:Carousell 的运营模式已是生产级、云原生,并且在客户和内部工作流两端越来越多由 AI 介入。已记录技术栈也意味着显著供应商集中,但它显然已超出轻量市场平台基础设施。[CE011, CE012, CE013, CE014, CE015, CE016]
| 层级 / 组件 | 作用 | 依赖 | 风险 |
|---|---|---|---|
| Gemini 多模态模型 | 生成 AI 辅助上架内容,并丰富发现体验 | Google Cloud | 模型质量和延迟可能影响转化 |
| Vertex AI + BigQuery | 对聊天数据做批量推理,快速迭代欺诈模型 | Google Cloud 数据栈 | 供应商集中和成本治理 |
| Compute Engine C3D | 迁移后承载计算密集型工作负载 | Google Cloud 基础设施 | 云依赖和容量经济性 |
| Cloudflare CDN / WAF / DNS / SSL 基础设施 | 加快资产分发并缓解攻击 | Cloudflare 网络边缘 | 外部平台依赖 |
| Marketplace app 和网页客户端 | 买家 / 卖家的主要界面 | 移动应用商店、浏览器、内部产品团队 | App 质量和发布节奏会影响信任 |
| 审核和客服运营 | 人工审核、纠纷处理和政策执行 | 内部团队和政策工具 | 人力强度和误报会侵蚀 UX |
架构表只覆盖公开证据能支持的内容;许多内部服务仍未披露。
[CE011, CE014, CE017, CE020, CE021, CE024]Carousell 的产品栈结合了消费者 UX、信任服务和云基础设施。
[CE003, CE011, CE014, CE020, CE029]5.3 差异化、工程信号与路线图
产品成熟度最可信的证明,不只是 Carousell 自己说了什么,而是其伙伴和公开工程界面展示了什么。GitHub 显示,公司有一个公开组织,仓库覆盖 Go、Swift、Kotlin、TypeScript、Python 和 shell 工具,包括围绕 grpc/protobuf 构建的 Orion 微服务框架。Google Cloud 还描述了一条路线图,从当前 AI 辅助发布体验延伸到低于 45 秒的商品发布时间、更丰富的发现用风格属性,以及可能缩短工程工作的内部智能体式工具。这些本身都不能证明硬技术护城河。更大竞争对手可以买入或复制许多通用 AI 能力。但它确实显示出一个严肃的产品和平台组织,能把基础设施、数据和流程改善转化为面向用户的速度和信任收益。Carousell 的真实边际优势看起来更偏运营和整合,而非纯算法。它也说明产品迭代是连续的,而不是偶发的;基础设施和机器学习变化会直接进入用户工作流。[CE022, CE023, CE031, CE032, CE033, CE034]
| 日期 / 阶段 | 功能 / 里程碑 | 状态 | 含义 | 来源 |
|---|---|---|---|---|
| FY25 / FY26 | AI 作为战略重点 | 已宣布 | 表明公司会继续围绕信任和效率投入产品 | Carousell Press |
| 当前生产环境 | AI 辅助发布 | 已上线并持续改进 | 核心工作流压缩已进入市场 | Google Cloud |
| 目标状态 | 上架时间低于 45 秒 | 探索 / 目标 | 进一步降低卖家摩擦,并提升供给质量 | Google Cloud |
| 31 Jan 2026 前 | Passkey / 生物识别登录和黑名单防护 | 已承诺 | 强化身份和账号接管防御 | MHA |
| 当前 | 面向内部任务探索智能体式编码 | 探索中 | 可能加快工程维护和升级 | Google Cloud |
路线图混合了已上线功能、明确承诺和已表述的探索目标。
[CE027, CE031, CE032, CE033]Carousell 最强的地方,似乎在产品流程与信任运营交汇处。
成熟度评分是有证据支持的序数判断,不是公司发布的评分表。
[CE013, CE022, CE034, CE035, CE037]5.4 信任、安全、合规与质量控制
信任和安全是核心产品功能,不是合规附带项。Carousell 自己的帮助中心资料描述了 AI 和机器学习检测、版主审核、争议提示、实时警告、手机验证,以及 2FA 或 SMS 重新验证。MHA 和其他独立报道还提供了更强证据:已验证账号被滥用于诈骗后,Carousell 承诺引入 passkeys 或生物识别登录、账号暂停控制和 Singpass 黑名单措施。MHA 交易安全页面也把用户徽章、安全支付、官方配送和记录留存描述为产品化防护。与此同时,制裁政策显示,公司还有一层处理不支持地区和禁止交易的合规层。合起来看,这些来源支持一个清晰结论:Carousell 已构建分层信任栈,覆盖界面、工作流、审核、身份和合规。未知的是这些系统的精确度,以及大规模运行它们的运营成本。换句话说,Carousell 交付的不只是市场平台应用,也是一套信任操作系统。[CE024, CE025, CE026, CE027, CE028, CE029]
| 控制项 / 指标 | 状态 | 范围 | 缺口 |
|---|---|---|---|
| AI/ML 诈骗检测 | 已启用 | 全平台欺诈和可疑活动检测 | 精确率和召回率未披露 |
| 链接、电子邮件和二维码拦截 | 已启用 | 聊天安全和钓鱼防护 | 误报率未知 |
| 手机号验证 | 已启用 | 发布商品或发起聊天时要求验证 | 各市场抗欺诈能力未知 |
| 2FA / SMS 重新验证 | 已启用 / 已强化 | 可疑登录和高风险流程 | 覆盖率和用户摩擦影响未知 |
| Passkeys / 生物识别登录 | 承诺 Jan 2026 前上线 | 更高安全性的认证路线图 | Jan 2026 后实施状态尚未公开 |
| Carousell Rights Owners Programme | 已启用 | 与品牌伙伴共建的反假货工作流 | 品牌覆盖和下架统计未公开 |
| 制裁 / 贸易合规筛查 | 已启用 | 不支持的国家、商品和交易 | 运营负担和例外处理未公开 |
信任控制覆盖界面警示、身份、审核和合规后台。
[CE024, CE025, CE026, CE027, CE028, CE029]关键产品结果取决于云、身份和审核系统协同运行。
[CE014, CE020, CE026, CE027, CE029]5.5 图表
06客户
6.1 客户分层与谁在使用 Carousell
Carousell 的客户群不是一个单一整体。公开证据至少支持四类不同用户:清理闲置物品的普通消费者、寻求低价的买家、垂直或专业商户,以及使用 CarouBiz 的企业卖家。品类宽度在这里很重要,因为每个分层的行为不同。家政服务、房产、汽车和奢侈品不只是不同产品;它们涉及不同信任需求、交易速度和变现潜力。Specialist Awards、认证企业徽章和 CarouBiz 条款都显示,Carousell 正有意识地在日常点对点使用之外培育更结构化的卖家类型。同时,应用商店语言和核心产品仍把平台定位为普通消费者可以低摩擦使用的工具。普通和专业参与者的混合,是 Carousell 扩大供给又不变成纯商户市场平台的核心。实际中,每个分层可能会对价格、信任、便利和速度的不同组合作出响应。[CU001, CU002, CU003, CU004, CU005, CU006]
| 客群 | 买家 / 用户 / 付款方 | 使用场景 | 规模信号 | 收入 / 战略价值 | 缺口 |
|---|---|---|---|---|---|
| 普通消费者 | 买家和卖家往往是同一个家庭用户 | 断舍离、捡便宜、本地交易 | Carousell 六个主要市场有数百万 app 用户 | 流动性引擎和品牌触达 | 未披露 GMV 或收入的具体占比 |
| 信任敏感型买家 | 手机、奢侈品、汽车等较高风险二手商品买家 | 需要验证、更安全的支付和配送 | 已有认证商品、Buyer Protection、门店和直卖项目 | 变现和复购潜力更高 | 未披露品类级复购率 |
| 专业商家 | 通过 Specialist Awards 或认证商家流程获得认可的商家 | 靠口碑和服务质量销售 | Specialist Awards 和认证商家徽章已公开 | 提升优质供给和客户信心 | 商家数量和 GMV 占比未知 |
| 企业卖家 | 使用 CarouBiz 订阅的企业 | 品牌建设、运营、重复发布 | CarouBiz 订阅条款已公开 | ARPU 和复用可能更高 | 套餐采用率和收入贡献未披露 |
客群是从公开产品和政策触点推断出来的,并非管理层明确给出的一张细分图。
[CU004, CU005, CU006, CU007, CU008, CU029]Carousell 希望把用户从随意发现推进到更可信、也更容易变现的重复流程。
[CU004, CU005, CU009, CU022, CU035]6.2 采用轨迹与可见增长信号
公开资料包对广义采用的支持强于对精确增长机制的解释。Carousell Group 称其服务七个市场的数千万用户,应用商店界面则提到数百万人使用应用清理闲置、赚钱和淘低价好物。公司目前没有在所审阅来源中发布精确 MAU,这限制了趋势精度,但它提供了几个支撑性采用信号。其中包括 29 家线下门店、强品类宽度,以及 Google Cloud 称已让上架商品数增加 20% 的 AI 驱动商品发布体验。DealStreetAsia 还报道,门店和官方以旧换新带来超过 20% 的 GMV 提升。这组信号说明,业务不是只守着一大批沉睡用户;它正在主动通过更低摩擦的商品发布和更高信任的线下触点,提高交易密度、信任和频次。[CU001, CU013, CU014, CU018, CU019, CU023]
| 指标 | 数值 | 日期 | 来源 | 置信度 | 含义 | 缺失分母 |
|---|---|---|---|---|---|---|
| 集团用户规模 | 数千万用户 | 2026 | Carousell Press / TechNode | 中 | 按区域消费者平台看,Carousell 规模已经很大 | 未披露各市场 MAU 精确拆分 |
| 应用商店评分 | iOS 页面 3.1K 条评分,4.7 / 5 | 2026 | Apple App Store | 中 | 可见的主流满意度信号较强 | 评分未按用户队列或时间新近度拆分 |
| 门店覆盖 | 29 家实体门店 | FY25 / 2026 | Carousell Press / DealStreetAsia | 高 | 线下触点已经成为客户体验的一部分 | 缺少单店流量或转化数据 |
| 门店 / 以旧换新提升 | GMV 提升 20%+ | FY25 | DealStreetAsia | 中 | 全渠道打法似乎加深了消费或转化 | 基线和测量窗口未公开 |
| 卖家供给代理指标 | 上架商品数增加 20% | 当前 | Google Cloud | 中 | AI 降低供给侧参与摩擦 | 新增上架与留存卖家之间没有公开关联 |
| 信任暴露 | 93%+ 活跃用户未遇到骗子 | 2025 年报告 | Carousell Help / Press | 高 | 信任环境可能已改善到足以守住使用 | 未按市场或品类拆分 |
采用指标在规模和信任上强,在留存队列和精确分母上弱。
[CU001, CU013, CU014, CU018, CU023, CU024]采用率改善有两条来源:用户规模扩大,可信交易路径摩擦降低。
[CU001, CU013, CU023, CU024, CU035]6.3 客户证明、信任信号与满意度证据
对消费者市场平台而言,客户证明更多来自围绕评价、信誉信号和真实交易防护的分层证据,而不是知名企业标识。Carousell 的评价帮助页显示,它有真实的双向评价系统,买卖双方会对彼此的交易质量打分,并留下可成为可见个人主页资产的称赞。Specialist Awards 和认证企业徽章增加了另一层商户证明,尤其面向意图更高的买家。信任栈还得到公开诈骗和验证指标强化:2025 年 99.96% 的交易无诈骗,超过 93% 用户未遇到诈骗者,并对可疑账号进行大规模执法。重要但书是,外部客户情绪并不完全干净。应用商店评分强劲,但评价聚合网站仍浮现安全和支持质量方面的警示信号。正确结论是,Carousell 有有意义的客户活动和信任证明,但满意度并非无摩擦。这个差异很重要,因为市场平台健康度依赖许多小信任线索复合,最终转化为再次交易意愿。[CU009, CU010, CU011, CU012, CU015, CU016]
| 客户 / 证明集 | 客群 | 部署 / 使用场景 | 生产环境还是试点 | 结果 | 局限 |
|---|---|---|---|---|---|
| iOS App Store 评价者 | 大众买家和卖家 | 一般买卖和服务使用 | 生产环境 | 评分 4.7/5,实时评分量很大 | 应用商店评价偏向活跃移动端用户 |
| Specialist Award 商家 | 意图更强的专业卖家 | 专业型或重视声誉的销售 | 生产环境 | 面向客户服务和专业能力的公开徽章 | 未披露 GMV 或复购 |
| 使用购买按钮并互评的买卖双方 | 可信交易用户 | 双边都有评分的结构化交易 | 生产环境 | 互评和称赞体系支撑信任闭环 | 无法揭示长期留存 |
| 奢侈品或认证品类用户 | 对信任要求更高的买家和寄售方 | 经过认证的奢侈品和直接销售流程 | 生产环境 | 认证和信心承诺支撑高端品类 | 结果细节更能说明信任,而不是频次 |
交易平台上的“具名证明”,更适合理解为可识别的证明界面和商家集合,而不是企业 logo 客户。
[CU007, CU009, CU011, CU014, CU028]客户验证在广泛活跃度和信任信号上最强,在精确留存证据上较弱。
矩阵是有证据支持的分析判断,不是公司口径。
[CU014, CU015, CU017, CU018, CU019, CU037]6.4 持久性、扩张与集中风险
持久性问题上,公开证据变薄。Carousell 可能的先落地再扩张路径不是经典企业增购,而是把用户迁移到更高信任、更高变现的工作流:Buy button、认证品类、垂直商户、Sell to Carousell 和门店。这些路径可以加深参与度、提高每活跃用户收入,尤其在手机、奢侈品和汽车中。客户集中风险也与企业软件结构不同。Carousell 不太可能依赖少数命名账号,但仍可能按地理、信任监管和卖家结构集中。新加坡的重要性不成比例,因为最可见的验证、Singpass 和反诈骗要求都在那里有记录。最重要的是,所审阅公开来源没有披露真实留存、流失、用户分群行为或专业卖家占比。因此,投资人看到的是好的规模故事、还不错的信任故事,以及不完整的持久性故事。这也意味着,管理层披露用户分群的价值会远高于又一个宽口径用户数标题。[CU023, CU025, CU028, CU030, CU031, CU032]
| 指标 | 数值 / 空值 | 分群 | 置信度 | 尽调追问 |
|---|---|---|---|---|
| 应用商店满意度 | iOS 评分 4.7/5 | 广泛移动端用户 | 中 | 与争议率和客服解决 CSAT 对照 |
| 评价闭环机制 | 30 天评价窗口、称赞、互评 | 买家和卖家 | 中 | 要求披露完成交易中产生评价的占比 |
| 外部情绪汇总 | 好坏参半 / 需要谨慎 | 独立评价读者 | 低 | 用应用内 CSAT 和投诉量校准 |
| 复购 / 重复发布率 | Null | 全部分群 | 低 | 按品类要求复购订单和重复发布队列 |
| GRR / NRR / 流失率 | Null | 全部分群 | 低 | 要求披露买家、商家和高活跃卖家的留存指标 |
| 信任调整后的韧性 | 缺失,但方向上改善 | 风险更高的品类 | 低 | 要求披露认证流程的复购和争议率 |
公开来源仍未解答多数韧性问题;本表直接保留这种不确定性。
[CU009, CU010, CU011, CU015, CU017, CU030]| 扩张驱动 | 集中度风险 | 影响 | 尽调路径 |
|---|---|---|---|
| 购买按钮和 Buyer Protection | 如果重视信任的买家仍转到站外,使用率可能保持低位 | 限制变现和留存提升 | 按品类和市场要求购买按钮渗透率 |
| 认证品类和门店 | 经济性可能集中在少数高价值品类 | 能提升 ARPU,但会压窄变现广度 | 要求披露手机、奢侈品和汽车的 GMV 与复购率拆分 |
| CarouBiz 和专业商家 | 专业卖家占比可能仍小 | 如果休闲供给占主导,扩张论据会变弱 | 要求披露商家层级收入组合和重复发布节奏 |
| 新加坡主导的验证和信任监管 | 客户旅程可能只有在新加坡最成熟 | 其他市场的信任产品成熟度可能落后 | 要求逐市场披露信任功能采用率 |
| 主应用覆盖六个市场 | 增长可能依赖一组有限的成熟市场 | 地理集中会限制增量增长 | 按市场要求活跃用户和收入 |
Carousell 的集中度更可能体现在市场和卖家组合,而不是少数具名账户。
[CU023, CU025, CU032, CU033, CU034, CU035]公开留存证据来自可见信号,而不是数字化留存队列。
本章明确保留留存缺口,因为没有找到公开的留存队列百分比。
[CU030, CU031, CU036, CU038]6.5 图表
07风险
7.1 监管与法律风险
监管压力对 Carousell 不是假设,已经在新加坡塑造产品要求。MHA 已直接按 Online Criminal Harms Act 框架评估该平台,并在已验证账号滥用持续后要求更强控制。仅这一点就把监管风险抬高到普通消费者平台担忧之上。随着 Carousell 在多个司法辖区处理敏感身份、支付、位置、聊天和争议数据,法律和政策复杂度也在扩张。制裁和 AML 政策显示,公司围绕地理范围和金融犯罪筛查还有另一层合规开销。这些都不意味着公司特别不合规,但意味着风险会累积:诈骗防范、身份验证、隐私和跨境合规在一个面向客户的产品里汇聚。监管机构可能比平台低成本适配更快地收紧要求,因此合规成本和用户摩擦都是实质下行通道。条款和禁售内容规则也显示,Carousell 必须在细粒度层面治理行为、商品发布和身份,交易量扩张时政策执行负担会加重。[CR001, CR002, CR003, CR004, CR005, CR006]
| 规则 / 案件 | 司法辖区 | 状态 | 可能性 | 严重性 | 缓释措施 | 剩余暴露 | 尽调路径 |
|---|---|---|---|---|---|---|---|
| OCHA 强化验证和诈骗控制 | 新加坡 | 已生效并在升级 | 高 | 高 | 通行密钥 / 生物识别、黑名单、账号冻结、验证 | 高 | 按月要求监管沟通记录和落地状态 |
| 隐私和敏感数据处理 | 多司法辖区 | 持续有效的现行义务 | 中 | 高 | 政策控制、身份检查、争议证据流程 | 中-高 | 审查 DPIA、留存期限和访问控制 |
| 制裁 / AML / 不支持国家控制 | 多司法辖区 | 现行政策 | 中 | 中 | 商品发布和访问筛查、账号封禁 | 中 | 审查交易筛查流程和例外日志 |
| 假货和 IP 执法 | 多市场 | 已在执行,但始终存在争议 | 中 | 中 | CROP 合作方下架、认证和审核 | 中 | 要求披露假货索赔量和解决时间 |
排序按投资论点受冲击的严重性,而不是法理条文。
[CR001, CR002, CR004, CR005, CR011, CR014]最高严重度风险集中在信任 / 监管和竞争,而不是某一单一法律事件。
严重度是有证据支持的序数判断,不是公司发布的控制框架。
[CR001, CR010, CR029, CR037, CR041]7.2 运营、质量与安全风险
Carousell 的运营风险,是开放市场平台达到实质规模后会遇到的经典问题:钓鱼、交易迁移到平台外、订金诈骗、假货暴露、账号接管,以及随着平台推动更多交易在平台内完成而上升的可靠性预期。好消息是,Carousell 似乎运行着分层缓释栈,包括机器学习检测、链接拦截、评价循环、官方配送、Buyer Protection、争议提示和审核。坏消息是,每一种保护都会花钱、增加工作流摩擦,或在某些边缘情形下仍会失效。没有公开事故日志或可用性披露能解答可用性问题,商户条款还明确排除了 Carousell 控制之外的停机。因此,风险不只是欺诈存在;而是公司可能不得不持续大额投入控制措施,仅仅为了守住当前信任水平,同时在诈骗数量上升时仍面对公众批评。[CR009, CR016, CR017, CR018, CR019, CR020]
| 失效模式 | 可能性 | 严重性 | 缓释成熟度 | 剩余暴露 | 未解缺口 |
|---|---|---|---|---|---|
| 站外钓鱼和押金诈骗 | 高 | 高 | 中-高 | 高 | 公开资料未披露欺诈损失负担 |
| 账号被接管,包括已验证账号 | 中-高 | 高 | 中 | 高 | 需要通行密钥 / 生物识别的落地证明 |
| 假货或商品质量争议 | 中 | 中-高 | 中 | 中 | 没有公开缺陷率或争议解决统计 |
| 服务停机或信任流程降级 | 中 | 中 | 低-中 | 中 | 没有公开可用性 SLA 或事故历史 |
| 门店 / 以旧换新运营错误 | 中 | 中 | 中 | 中 | 没有公开门店质量或回本指标 |
最强的控制措施看得见;最强的运营成本证据看不见。
[CR016, CR017, CR018, CR019, CR020, CR024]信任和监管失灵会快速传导到客户行为、利润率和估值。
[CR002, CR010, CR018, CR025, CR041, CR042]7.3 依赖、财务与市场结构风险
Carousell 依赖复杂的外部栈。Google Cloud 位于 AI、数据和计算工作流内部,Cloudflare 支撑核心流量处理和攻击缓释。应用商店、支付系统、配送伙伴和验证通道也调解客户旅程的一部分。在温和市场中,这可以管理;但外部市场远非温和。区域电商和社交电商现在由少数规模平台主导,Shopee 和 TikTok Shop 对需求形成和卖家经济性尤其有强拉力。对 Carousell 而言,风险既由伙伴驱动,也由市场结构驱动:云或平台问题会伤害运营,竞争环境加剧则会迫使更高信任支出、更低抽佣率目标,或更慢走向净盈利。EBITDA 转正确实存在,但没有移除竞争强度和仍不完整的公开单位经济所带来的模式风险。换句话说,Carousell 已不再是在一个平静的分类信息利基中竞争;它是在更严酷的区域电商栈中承销信任和交易质量。[CR021, CR022, CR023, CR025, CR026, CR028]
| 依赖项 | 交易对手 | 角色 | 集中度 | 失效情景 | 严重性 | 缓释措施 | 剩余暴露 |
|---|---|---|---|---|---|---|---|
| 核心 AI / 数据云 | Google Cloud | 商品发布 AI、欺诈推断、数据处理、算力 | 高 | 服务中断、价格冲击或政策变化伤及核心工作流 | 高 | 多工具布局和内部工程深度可能减轻部分切换痛点 | 高 |
| 边缘安全和流量处理 | Cloudflare | CDN、WAF、DNS、SSL、DDoS 防护 | 高 | 流量峰值或攻击处理能力下降 | 中-高 | Cloudflare 集成已在峰值中经受实战检验 | 中 |
| 应用分发 | Apple / Google | 用户获取和更新渠道 | 高 | 政策或排名变化影响安装量和发版节奏 | 中 | 跨平台存在和网页端触点 | 中 |
| 社交和交易平台替代品 | TikTok Shop / Facebook Marketplace / Shopee 同类 | 需求截获和卖家注意力 | 高 | 发现流量和 GMV 从 Carousell 外流 | 高 | 品类聚焦和信任流程 | 高 |
| 支付和配送通道 | 第三方支付和物流合作伙伴 | 促成可信结账 | 中 | 糟糕的承运商或支付体验伤害信任 | 中 | 官方配送和应用内记录有帮助 | 中 |
合作伙伴风险和竞争风险相互重叠,因为平台依赖和市场份额集中会彼此强化。
[CR021, CR022, CR023, CR029, CR030, CR031]关键依赖位于基础设施、分发和需求捕获的交汇处。
[CR021, CR022, CR023, CR029, CR032, CR033]7.4 人员、执行与投资逻辑破裂触发点
执行风险仍是最重要的投资风险之一,因为战略本身对运营要求很高。Carousell 正同时提高信任、让卖家专业化、运营门店和以旧换新、整合收购,并改善盈利。公司已经证明,校准失误会带来痛苦修正,包括 2022 年和 2024 年的裁员。这不证明管理层弱;但证明战略几乎没有给粗糙执行留下空间。创始人 CEO 仍是资本配置、盈利叙事和战略方向的公开面孔,增加了关键人物敏感性。因此,投资人应少问风险是否存在,多看风险是否被转化为可监控控制。若信任控制持续昂贵却无效,若品类运营赚不回成本,或竞争迫使 Carousell 回到重补贴行为,投资逻辑就会破裂。[CR027, CR037, CR038, CR039, CR040, CR041]
| 角色 / 职能 | 依赖或缺口 | 可能性 | 严重性 | 缓释措施 | 尽调路径 |
|---|---|---|---|---|---|
| 创始人 CEO 领导 | 战略和资本配置集中在公开可见的创始人领导层 | 中 | 高 | 已有更宽的高管梯队,但公开能见度仍集中 | 要求披露继任安排和授权运营结构 |
| 信任和审核团队 | 即使有 AI 工具,仍需要人工监督 | 中 | 中-高 | AI + 政策 + 审核员组合 | 要求披露工作量、积压和误报跟踪 |
| 门店 / 品类运营 | 检测、质量和全渠道执行复杂度 | 中 | 高 | 标准化流程和品类专业化 | 要求披露 QA 率和品类级门店运营手册 |
| 收购整合 | 已收购品牌和运营组合 | 中 | 中-高 | 品牌级专业化可能有帮助 | 要求整合评分卡和重叠经济性 |
| 成本纪律 | 过度投资后曾裁员 | 中 | 中-高 | EBITDA 里程碑显示纪律有所改善 | 要求披露人员计划和招聘效率 |
执行风险并不抽象;管理层已经不得不公开纠偏。
[CR027, CR037, CR038, CR039]| 风险 | 可监测触发项 | 阈值 / 事件 | 行动含义 |
|---|---|---|---|
| 信任控制失效 | 新防护上线后监管干预增加 | 即便 2026 年 1 月控制措施落地,仍被要求再次升级 | 暂停乐观预期,重估信任经济性 |
| 欺诈负担压垮经济性 | 欺诈损失或争议成本增长快于交易增长 | 管理层无法证明争议成本曲线改善 | 假设 EBITDA 质量更低、准备金需求更高 |
| 门店 / 以旧换新运营不及预期 | 没有回本或转化提升证据 | 门店不成护城河,反而拖累流量或利润率 | 因运营强度下调估值 |
| 竞争压缩模式 | 增长放缓,同时抽佣率或促销条件恶化 | Carousell 被迫采取高补贴回应 | 根据现金影响转向不通过 / 强不通过 |
| 管理层执行失速 | 又一次大规模裁员,或在经济性未改善下重置战略 | 新投资周期后反复纠偏 | 视为运营纪律上的投资论点破裂 |
否决标准聚焦可观察的运营证据,而不是叙事层面的不适。
[CR025, CR028, CR034, CR041, CR042]7.5 图表
08估值
8.1 估值锚点与业务质量变化
Carousell 估值的起点仍是 2021 年 9 月 US$1.1B 的独角兽轮,因为之后没有更干净的公开融资标记取代它。这个锚点已经旧了,但并非无关紧要。几年下来,公司从仍在亏损但保持增长的市场平台,变成 FY25 收入 US$141M、EBITDA 为正的业务。因此,运营画像比估值时间戳本身更好。不过,投资人不能因为业务质量改善就简单上调估值。缺的东西仍然是公开 GMV、分品类抽佣率、可持续净利润,以及二手电商门店和专业垂类能赚到有吸引力回报的证据。正确框架不是「旧估值等于错估值」,而是「旧估值现在必须拿更强的基本面和仍然不小的不确定性重新判断」。投资人也要记住,私募市场标记可以在基本面改善时保持不变;这种动态本身不说明当前价格错了,但确实让情景分析比标题式怀旧更重要。[CV001, CV002, CV003, CV004, CV005, CV006]
| 锚点 | 数值 | 日期 | 解读 | 置信度 | 声明引用 |
|---|---|---|---|---|---|
| 上一次公开融资估值 | US$1.1B | 2021-09 | 仍是最干净的公开定价标记 | 高 | CV001 |
| FY25 收入 | US$141M | 2026-07 | 做倍数分析的当前规模基数 | 高 | CV002;CV003 |
| FY25 EBITDA 状态 | 为正 | 2026-07 | 经营质量较往年改善 | 高 | CV003 |
| FY24 净亏损 | US$33.2M | 2025-07 | 显示改善起点仍是亏损基准 | 中 | CV004 |
| 累计融资总额 | 约 US$300M+ 区间 | 2026-07 | 缓解融资压力,但不能消除估值争议 | 中 | CV006 |
Carousell 最近一次公开价格发现仍停留在 2021 年。
[CV001, CV002, CV003, CV004, CV006]| 因素 | 方向 | 幅度 | 逻辑 |
|---|---|---|---|
| EBITDA 里程碑 | 溢价 | 中 | 相较历史亏损画像,经营质量更好 |
| 二手循环商业收入占比 45% | 溢价 | 中 | 比传统分类广告更可防守,也更由交易驱动 |
| 门店和垂类专家 | 溢价 / 折价 | 中 | 既可能加深护城河,也可能拖累运营 |
| 东南亚竞争集中 | 折价 | 高 | Shopee / TikTok / Lazada 挤压流量和卖家经济性 |
| GMV 和单位经济性披露缺口 | 折价 | 高 | 限制高溢价定价的信心 |
| 私募稀缺性与上市可比流动性 | 溢价 / 折价 | 中 | 私募赢家可拿溢价,公开市场估值标记会约束定价 |
部分因素双向影响:战略野心既可能造护城河,也可能变成成本。
[CV007, CV008, CV021, CV022, CV023, CV025]Carousell 介于披露透明的上市转售可比公司与高端私营二手电商赢家之间。
坐标轴按公开披露归一化,只作方向性参考,并非精确财务评分。
[CV003, CV010, CV013, CV016, CV018]8.2 可比公司组与倍数框架
没有任何一个可比公司完全贴合。Vinted 是最好的高端私募可比,因为它同时具备二手行为、平台流动性、内嵌物流和支付,以及可见的盈利能力。Back Market 可以作为更高信任度的二手电商可比,服务层和可信供给支撑更厚的经济性。ThredUp 则是透明度最好的公开市场可比,尽管公开市场在结构上会给它相对私募战略赢家更低的折价。这些同业说明,如果投资人相信 Carousell 交易驱动的二手电商策略正在奏效,它的交易倍数应低于最强私募龙头,但高于商品化或纯公开市场标记。难点在于,Carousell 规模比高端同业小,同时运营野心又比轻型市场平台更重。这个组合要求估值工作保持克制,而不是追求极致。地理位置也重要:Carousell 竞争的是变化更快的东南亚商业栈,而 Vinted 受益于更广的欧洲足迹,Back Market 受益于更强的翻新供给标准化。[CV009, CV010, CV011, CV012, CV013, CV014]
| 公司 | 商业模式 | 最新收入基数 | 盈利信号 | 最新估值 / 市值 | 隐含收入倍数 |
|---|---|---|---|---|---|
| Carousell | 东南亚交易平台 / 二手循环商业 | US$141M FY25 | FY25 EBITDA 为正 | 历史公开估值 US$1.1B | ~7.8x |
| Vinted | 欧洲 C2C 二手交易平台 | €1.1B FY25 | €62M 净利润;€151M 调整后 EBITDA | 二级市场估值 €8B | ~7.3x |
| Back Market | 翻新电子产品交易平台 | 2025 年估计 US$495M | EBITDA 打平 | 估计估值 US$5.7B | ~11.5x |
| ThredUp | 上市转售交易平台 | 2026 年指引 US$351M-356M | 调整后 EBITDA 利润率指引约 6.1% | 市值参考约 US$775M | ~2.2x |
币种不同、公私市场口径不同,精确倍数不能完全对齐;本表只看方向。
[CV002, CV010, CV011, CV013, CV016, CV017]| 框架 | 收入倍数 | 基于 FY25 收入的隐含价值 | 设置原因 | 成立前提 |
|---|---|---|---|---|
| 类上市公司下行情景 | 3x-4x | US$423M-US$564M | 上市可比折价、披露缺口、竞争压力 | 投资人优先看透明度和有流动性的可比标的 |
| 平衡的私募区间 | 5x-7x | US$705M-US$987M | EBITDA 里程碑和战略可选性,但风险折价仍在 | 信任改善和品类经济性守得住 |
| 高溢价二手循环商业区间 | 7x-9x | US$987M-US$1.27B | 稀缺区域赢家叙事 | Carousell 表现更像 Vinted / Back Market,而不是上市可比公司 |
区间是启发式估算,不是 DCF 输出。
[CV027, CV028, CV029, CV030, CV032]Carousell 历史隐含倍数低于 Back Market,接近 Vinted,远高于已上市的 ThredUp。
私营和上市可比对象币种不一,这里只作方向性参照。
[CV002, CV011, CV013, CV017, CV018]8.3 估值区间与情景分析
第一轮投资判断,用简单倍数框架就足够。若市场坚持透明度、上市可比折价和未解决的竞争风险,偏公开市场的下行情景可以支撑约 3x 到 4x 收入的区间。较均衡的私募市场区间约为 5x 到 7x 收入,反映 EBITDA 转正、品类扩张和持续改善的信任栈,但仍对规模和披露缺口打折。只有当投资人把 Carousell 视为稀缺区域赢家,并看到门店、专家型垂类和 AI 强化信任工具是在加深护城河而不是增加成本时,约 7x 到 9x 收入的高端二手电商区间才站得住。这个练习的结论是,US$1.1B 已不再明显虚高,但仍接近 2026 年可辩护区间的乐观端。实际操作中,这是谈判区间,不是单一真理;投资人愿意在区间内哪个位置交易,应由尽调质量决定。[CV021, CV022, CV023, CV024, CV025, CV026]
| 情景 | 价值区间 | 主要假设 | 触发条件 / 支撑 |
|---|---|---|---|
| 悲观 | US$550M-US$800M | 增长放缓,信任投入上升,上市可比压力占主导 | FY26 证据偏弱或竞争恶化时触发 |
| 基准 | US$900M-US$1.15B | EBITDA 维持为正,二手循环商业 mix 改善,无重大信任倒退 | 持续执行且披露适度时支撑 |
| 乐观 | US$1.2B-US$1.4B | 门店和垂类专家证明有产出,AI 拉动转化,护城河加深 | 需要证明单位经济性和信任水平同步改善 |
基准情景围绕当前公开独角兽锚点展开。
[CV032, CV033, CV034]当前独角兽估值落在 2026 年可支撑区间的上半段。
用 FY25 收入乘以经验区间直接测算。
[CV027, CV028, CV029, CV032]8.4 投资立场
最站得住的立场,是有条件通过,并给予贴近市场到略高于市场的估值判断。Carousell 通过真实收入规模、EBITDA 里程碑、不断提升的二手电商占比,以及相比 2021 年更具结构差异化的策略,配得上这个更好的态度。不过,公司还没有资格拿到接近最佳私募二手电商龙头的无条件溢价标记,因为它仍有披露缺口、区域竞争压力、信任成本不确定性,以及门店和垂直运营带来的执行复杂度。只要管理层能证明耐久性,而不只是证明动能,投资人就可以对当前估值保持开放。下一段上行更少取决于叙事,更取决于 FY26 能否把 EBITDA 转正转化为可重复、风险调整后质量更高的运营表现。在这些证据出现之前,纪律应来自条款、治理要求和基于里程碑的信念,而不是假装某个点估值足够精确。[CV035, CV036, CV037, CV038, CV039, CV040]
| 问题 | 当前答案 | 信心 | 含义 |
|---|---|---|---|
| 按最后公开估值看,Carousell 是否明显高估? | 不,未明显高估 | 中 | 不要仅因 2021 年头部估值就否决 |
| 相对风险,Carousell 是否明显便宜? | 否 | 中 | 条款和治理要求都要守纪律 |
| 经营进展是否足以支持认真接触? | 是 | 高 | 有条件通过,而不是否决 |
| 什么会提高信心? | 可持续 EBITDA、门店产出、更清晰的信任经济性 | 中 | 可支持更明确通过,并收窄溢价区间 |
| 什么会降低信心? | 增长放缓,同时信任成本上升或收入 mix 质量走弱 | 中 | 会把估值推向上市可比区间 |
矩阵把估值工作落到实际承销立场上。
[CV030, CV031, CV037, CV038, CV039, CV040]战略和财务改善撑起估值,但竞争压力和披露缺口抵消了大部分溢价逻辑。
单元格概括估值压力方向,不给出点估计。
[CV003, CV025, CV026, CV030, CV037]8.5 图表
免责声明
本报告是基于公开证据的尽调快照,不构成投资建议。关键财务、法律、技术和合同事实仍未公开;作出任何投资决定前,应直接向管理层和原始文件核验。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Carousell Group says it was founded in August 2012 in Singapore. | 高 | SO001, SO024 |
| CO002 | Public company materials identify Siu Rui Quek, Marcus Tan, and Lucas Ngoo as the founding team. | 中 | SO006, SO019 |
| CO003 | Carousell's 2026 fact sheet lists a leading presence in Singapore, Hong Kong, Malaysia, Indonesia, the Philippines, Taiwan, and Vietnam. | 高 | SO001, SO009 |
| CO004 | Carousell's 2021 unicorn announcement described operations across eight markets under the Carousell, Mudah.my, Cho Tot, and OneKyat brands. | 高 | SO024, SO025 |
| CO005 | Current group brands disclosed by Carousell include Carousell, Cho Tot, Mudah.my, Laku6, REFASH, LuxLexicon, OneShift, and Revo Financial. | 中 | SO001 |
| CO006 | Carousell's mission statement is to make secondhand the first choice for everyone. | 高 | SO001, SO024 |
| CO007 | Carousell reported FY25 revenue of $141 million, up 18% year on year. | 高 | SO002, SO003 |
| CO008 | Carousell said it reached positive EBITDA in FY25. | 高 | SO002, SO003 |
| CO009 | Carousell said recommerce revenue grew 40% year on year and represented 45% of total FY25 revenue. | 高 | SO002, SO003 |
| CO010 | By the end of FY25 Carousell operated 29 physical stores across Singapore, Hong Kong, Malaysia, and Indonesia. | 高 | SO002, SO009 |
| CO011 | Carousell said those stores drove more than 20% GMV uplift in its luxury, mobile phones, and fashion businesses. | 高 | SO002, SO003 |
| CO012 | Carousell said AI listing tools can help sellers create listings in as little as three seconds. | 高 | SO002, SO009 |
| CO013 | Carousell said more than 99% of transactions across its markets are completed without a scam incident. | 高 | SO002, SO009 |
| CO014 | Online Marketplaces reported FY24 revenue of $119.3 million. | 中 | SO005, SO020 |
| CO015 | Online Marketplaces reported FY24 net losses of $33.2 million, down 13% year on year. | 中 | SO005 |
| CO016 | Online Marketplaces reported FY24 cash and cash equivalents of $50.5 million. | 中 | SO005 |
| CO017 | Online Marketplaces reported FY24 classifieds advertising revenue of $52.8 million. | 中 | SO005 |
| CO018 | Online Marketplaces reported FY24 revenue split roughly between recommerce at $53 million, classifieds advertising at $52.8 million, and other revenue at $13.5 million. | 中 | SO005 |
| CO019 | Carousell made 76 employees redundant in December 2024, roughly 7% of workforce according to industry coverage. | 中 | SO005 |
| CO020 | Carousell CFO JJ Ang said the group sharpened investment in its core classifieds platform and iterated recommerce toward higher-margin business models. | 中 | SO005 |
| CO021 | Carousell raised US$100 million in September 2021 in a Series E round led by STIC Investments. | 高 | SO024, SO025 |
| CO022 | Carousell's September 2021 round valued the company at US$1.1 billion. | 高 | SO024, SO025 |
| CO023 | STIC managing director Jason Cho joined Carousell's board as part of the 2021 funding round. | 高 | SO024, SO025 |
| CO024 | Carousell's fact sheet describes its user base as tens of millions. | 高 | SO001, SO020 |
| CO025 | Carousell's fact sheet describes team size as 800+ teammates. | 中 | SO001 |
| CO026 | Dealroom's 2026 public preview maps 1,239 employees and team presence in 16 countries. | 低 | SO008 |
| CO027 | Public headcount figures conflict, so exact current employee count is not well corroborated from public sources. | 中 | SO001, SO008, SO020 |
| CO028 | Major investors named across public sources include Telenor Group, Rakuten Ventures, Naver, STIC Investments, 500 Global, Peak XV Partners, and EDBI. | 高 | SO001, SO024, SO025 |
| CO029 | Dealroom's public profile estimates Telenor Group owns 26.2% of Carousell's cap table. | 低 | SO008 |
| CO030 | CEO Quek Siu Rui said Carousell remained well capitalised in 2025 and that an IPO was only one of several possible future financing options. | 中 | SO004 |
| CO031 | The Straits Times reported earlier SPAC listing discussions ended amid macroeconomic uncertainty and valuation concerns. | 中 | SO004 |
| CO032 | Carousell launched a new headquarters at Block 79 Ayer Rajah Crescent in 2025. | 中 | SO004 |
| CO033 | The Straits Times reported Carousell earlier operated from Block 71, then Keppel Towers, and later Rochester Park before moving to Block 79. | 中 | SO004 |
| CO034 | Public coverage says Carousell's recent acquisition set includes REFASH, Ox Street, and Laku6. | 中 | SO004, SO019 |
| CO035 | Google Cloud said Carousell's AI-driven sell form lifted items listed per seller by 20% while reducing median listing time to under two minutes. | 中 | SO010 |
| CO036 | Google Cloud said Carousell now automates 90% of translation needs and eliminated a key compute cluster's on-call burden after infrastructure migration. | 中 | SO010 |
| CO037 | Cloudflare's 2020 case study said Carousell had more than 250 million listings and served roughly 1 PB of images per month. | 中 | SO011 |
| CO038 | Cloudflare's 2020 case study said flash sales can attract more than three times normal traffic. | 中 | SO011 |
| CM001 | Carousell participates in the broader Southeast Asian e-commerce and social-commerce economy, but its direct arena is narrower: secondhand consumer goods, related seller services, and trust-led classifieds. | 中 | SM001, SM002 |
| CM002 | Carousell's disclosed transaction categories span cars, luxury, mobile phones, fashion, home and living, property, and services. | 高 | SM001, SM012 |
| CM003 | Carousell's management now frames the company as a transaction-led recommerce platform rather than only a classifieds venue. | 高 | SM002, SM003 |
| CM004 | UnivDatos estimated the Southeast Asia e-commerce market at US$201.92 billion in 2024. | 中 | SM010 |
| CM005 | AFFiNCO estimated Southeast Asian e-commerce GMV at US$234 billion in 2026, up 18% year over year. | 低 | SM009 |
| CM006 | Sellercraft estimated regional e-commerce GMV would reach about US$230 billion by 2026. | 低 | SM007 |
| CM007 | Hashmeta cited about US$139 billion of Southeast Asian e-commerce GMV in 2023. | 中 | SM008 |
| CM008 | Statista's Southeast Asia topic page lists regional e-commerce market value at US$181 billion and internet-economy size at US$263 billion. | 中 | SM006 |
| CM009 | Indonesia is the region's largest e-commerce market by value in nearly every reviewed source. | 高 | SM006, SM009, SM010 |
| CM010 | AFFiNCO projects Indonesia at roughly US$121 billion of 2026 e-commerce GMV. | 低 | SM009 |
| CM011 | Vietnam is widely described as one of the fastest-growing Southeast Asian e-commerce markets. | 高 | SM006, SM009, SM010 |
| CM012 | AFFiNCO projects Vietnam's 2026 e-commerce GMV at about US$28 billion with 29% growth. | 低 | SM009 |
| CM013 | Singapore is smaller by GMV than Indonesia or Vietnam but has the region's highest order values or spending per shopper in reviewed sources. | 中 | SM007, SM009 |
| CM014 | Sellercraft says Singapore online spending exceeds US$1,200 per shopper annually. | 中 | SM007 |
| CM015 | AFFiNCO puts Singapore's 2026 e-commerce GMV at about US$7.2 billion with the region's highest average order value. | 低 | SM009 |
| CM016 | Mobile devices account for the large majority of Southeast Asian e-commerce transactions according to multiple sources. | 中 | SM007, SM009 |
| CM017 | AFFiNCO estimated mobile accounts for 78% of all e-commerce transactions in 2026. | 低 | SM009 |
| CM018 | Sellercraft estimated over 90% of online transactions in the region are completed through smartphones. | 低 | SM007 |
| CM019 | Social commerce is not a marginal channel in Southeast Asia; it is a major share of regional e-commerce value. | 高 | SM006, SM009 |
| CM020 | AFFiNCO estimated social commerce at 27% of regional e-commerce and cited 47% of shoppers buying during or after live-stream sessions. | 低 | SM009 |
| CM021 | Statista says TikTok Shop has already become a major regional competitor alongside Shopee and Lazada. | 中 | SM006 |
| CM022 | Carousell's market is therefore shaped not just by classifieds peers but by social-commerce discovery giants and general e-commerce platforms. | 中 | SM006, SM014, SM015, SM017 |
| CM023 | Electronics are one of the most important online categories across Southeast Asia. | 中 | SM005, SM007, SM010 |
| CM024 | Sellercraft says electronics account for roughly 20-30% of total regional e-commerce GMV. | 中 | SM007 |
| CM025 | Fashion is another leading category and a core beneficiary of social-commerce behavior in the region. | 高 | SM006, SM007 |
| CM026 | Carousell is exposed to both categories through fashion, luxury, and mobile phones. | 中 | SM001, SM002, SM012 |
| CM027 | Trust-sensitive categories such as luxury, mobile devices, and cars create room for Carousell's certified and sell-to-platform offerings. | 中 | SM002, SM013 |
| CM028 | Price comparison and review behavior are structurally important in Southeast Asian online shopping. | 中 | SM009, SM008 |
| CM029 | AFFiNCO says 83% of shoppers compare prices across platforms and 89% read reviews before buying. | 低 | SM009 |
| CM030 | Hashmeta similarly highlights reviews, influencer recommendations, and live-stream sessions as material conversion drivers. | 中 | SM008 |
| CM031 | Digital wallets are now a default expectation in much of Southeast Asia. | 中 | SM007, SM009, SM010 |
| CM032 | AFFiNCO estimates digital wallets handle 72% of regional e-commerce transactions while cash-on-delivery still persists in parts of the region. | 低 | SM009 |
| CM033 | Carousell's integrated payment, Buyer Protection, and official delivery layers are therefore directly aligned with regional buying behavior. | 中 | SM001, SM013 |
| CM034 | Omnichannel trust remains valuable because offline retail still captures most spending in major Southeast Asian markets. | 中 | SM007 |
| CM035 | Sellercraft says offline retail still captures roughly 80-85% of sales, which supports Carousell's physical-store strategy in higher-value categories. | 中 | SM007, SM002 |
| CM036 | Carousell lacks coverage in Thailand and other non-core markets, so its true serviceable market is smaller than headline SEA e-commerce TAM estimates. | 中 | SM001, SM010 |
| CM037 | Facebook Marketplace, eBay, Mercari, OLX, and other multi-category platforms remain viable substitutes for at least part of the same user job. | 高 | SM014, SM015, SM016, SM017, SM023 |
| CM038 | Carousell's strongest structural wedge is not owning all of SEA e-commerce, but winning trusted secondhand transactions where convenience and inspection matter. | 中 | SM002, SM003 |
| CP001 | Carousell competes in at least four competitor classes: social/discovery substitutes, global marketplaces, regional classifieds, and vertical resale specialists. | 高 | SP004, SP005, SP007, SP012 |
| CP002 | Facebook Marketplace is a major substitute because it bundles social identity, discovery, and local listing liquidity. | 高 | SP004, SP013 |
| CP003 | eBay is a major substitute because it spans many of the same categories and adds mature seller tooling and buyer trust programs. | 高 | SP005, SP017 |
| CP004 | Mercari competes most directly with Carousell on the “easy recommerce” narrative rather than classic classifieds. | 中 | SP006, SP003 |
| CP005 | OLX, Quikr, OfferUp, Craigslist, and Wallapop show the enduring strength of local-listing and classifieds alternatives around the world. | 高 | SP007, SP008, SP009, SP010, SP011, SP026 |
| CP006 | Depop is a narrower but relevant specialist competitor in fashion-led resale and discovery. | 中 | SP012, SP025 |
| CP007 | Third-party mapping sources consistently include eBay, OfferUp, Craigslist, Mercari, and Quikr among Carousell's closest visible web or business-model peers. | 中 | SP013, SP014, SP015 |
| CP008 | Carousell's clearest differentiation versus generic competitors is its move toward trusted, transaction-led recommerce with stores, certification, and safer payments. | 高 | SP002, SP003, SP018, SP019 |
| CP009 | Carousell has stronger regulatory trust signaling in Singapore than many lightweight classifieds substitutes because it publicly participates in enhanced marketplace verification efforts. | 高 | SP022, SP021 |
| CP010 | Carousell also has meaningful technical depth through Google Cloud and Cloudflare partnerships, which is relevant to AI, performance, and abuse prevention versus smaller classifieds peers. | 高 | SP023, SP024 |
| CP011 | Carousell remains weaker than the largest horizontal platforms on absolute liquidity and cross-border demand. | 中 | SP005, SP013, SP016 |
| CP012 | Facebook Marketplace and Craigslist can win low-friction local resale without needing deep transaction infrastructure, which limits how much Carousell can monetize pure listings. | 中 | SP004, SP010 |
| CP013 | eBay and Mercari demonstrate that seller tooling, trust layers, and repeat listing behavior are important determinants of durable marketplace economics. | 中 | SP006, SP017 |
| CP014 | Carousell's buyer protection and certified listing features give it a stronger offering than bare-listing substitutes in higher-risk categories. | 高 | SP018, SP019 |
| CP015 | Carousell's physical-store footprint is a competitive advantage in categories where inspection and instant sale matter more than broad assortment. | 中 | SP002, SP003 |
| CP016 | The strongest likely entrants are not new startups but larger social-commerce and ecommerce ecosystems that add local resale or trade-in flows. | 中 | SP004, SP005, SP006, SP027 |
| CP017 | Multi-homing is structurally easy in consumer marketplaces because the same seller can often cross-list inventory across several apps with low switching cost. | 中 | SP004, SP005, SP007 |
| CP018 | That low switching cost makes distribution, trust signals, and liquidity density more important than feature parity alone. | 中 | SP013, SP018, SP022 |
| CP019 | Carousell's moat is therefore conditional, not absolute: it is stronger in high-trust secondhand flows than in generic peer-to-peer listing. | 中 | SP002, SP003, SP022 |
| CP020 | Competitor pricing and packaging are often opaque on public surfaces, especially for classified boosts and seller subscriptions. | 中 | SP004, SP007, SP008 |
| CP021 | Because pricing is opaque, capability comparison is more reliable than headline list-price comparison for this chapter. | 中 | SP020, SP017 |
| CP022 | Carousell appears more feature-complete than Craigslist and bare listing substitutes on payments, verification, and safety content. | 中 | SP010, SP019, SP021 |
| CP023 | Carousell appears less globally liquid than eBay and less socially embedded than Facebook Marketplace. | 中 | SP004, SP005, SP013 |
| CP024 | Mercari is a better strategic analogue than Craigslist because both aim to simplify secondhand transactions rather than only host listings. | 中 | SP006, SP003 |
| CP025 | OLX and Quikr highlight how classifieds can persist, but they do not by themselves prove strong trust-led monetization. | 中 | SP007, SP008 |
| CP026 | Carousell's AI roadmap may strengthen ranking, trust, and service quality, but it is unlikely to create a durable moat by itself because larger platforms can match AI features. | 中 | SP002, SP023, SP005 |
| CP027 | Regulatory trust posture may be a more defensible wedge than AI features alone in Singapore and similar markets. | 中 | SP022, SP021 |
| CP028 | Carousell faces commoditization risk if users treat it as one of many interchangeable listing destinations. | 中 | SP013, SP017 |
| CP029 | Carousell faces displacement risk if TikTok-style discovery or social graph distribution wins more of the secondhand shopping journey before intent reaches a marketplace app. | 中 | SP004, SP025 |
| CP030 | Physical trust operations increase differentiation but may also increase execution complexity relative to fully digital competitors. | 中 | SP002, SP024 |
| CP031 | Smaller classifieds peers can still pressure Carousell on free listing behavior even if they lack Carousell's trust stack. | 中 | SP007, SP010 |
| CP032 | Carousell's combination of local-market focus, trust layers, and recommerce services is differentiated enough to support a “pass with moat caveats” stance. | 中 | SP002, SP003, SP022 |
| CP033 | Carousell's strongest competitor is whichever alternative best matches the category: Facebook for casual local resale, eBay for structured higher-scale resale, and specialists for verticals like fashion. | 中 | SP004, SP005, SP012 |
| CP034 | No reviewed source supports a claim that Carousell has winner-take-all control over resale in its markets. | 中 | SP013, SP015, SP016 |
| CP035 | The competitive question is not whether substitutes exist; it is whether Carousell can earn superior trust-adjusted take rates in the categories it serves. | 中 | SP002, SP019, SP022 |
| CI001 | Carousell reported US$141 million of FY25 revenue, up 18% year on year. | 高 | SI001, SI002, SI003 |
| CI002 | Carousell achieved positive EBITDA for the first time in FY25. | 高 | SI001, SI002, SI003 |
| CI003 | Recommerce contributed 45% of total FY25 revenue and grew nearly 40% year on year. | 高 | SI001, SI002, SI003 |
| CI004 | FY24 revenue was US$119.3 million, up 2.9% year on year. | 中 | SI004 |
| CI005 | FY24 net loss narrowed to US$33.2 million from the prior year. | 中 | SI004 |
| CI006 | FY24 cash and cash equivalents were US$50.5 million. | 中 | SI004 |
| CI007 | FY24 revenue split publicly disclosed by Online Marketplaces was roughly US$53 million recommerce, US$52.8 million classifieds advertising, and US$13.5 million other revenue. | 中 | SI004 |
| CI008 | The CFO said Carousell was iterating recommerce toward higher-margin business models. | 中 | SI004 |
| CI009 | Management told The Straits Times in 2023 that Carousell was well capitalised and not dependent on an IPO as the only route to raise capital. | 中 | SI005 |
| CI010 | Carousell raised roughly US$376 million across nine funding rounds according to Tracxn and The Company Check. | 高 | SI006, SI007 |
| CI011 | The latest visible external financing event in reviewed sources was a US$2.8 million September 2023 Series E extension led by GoTo Group. | 中 | SI006 |
| CI012 | The last clearly published post-money valuation anchor remains the US$1.1 billion September 2021 Series E round. | 高 | SI006, SI014, SI019 |
| CI013 | Carousell monetizes through at least three public streams: classifieds advertising, recommerce transactions, and other seller or service revenue. | 中 | SI004, SI011, SI013 |
| CI014 | Public app-store surfaces also show monetization via CarouBiz subscriptions, listing bumps, and coins, but they do not reveal realized take rates. | 中 | SI012, SI013 |
| CI015 | CarouBiz Lite is publicly listed at US$14.99 in the Canadian App Store. | 中 | SI012 |
| CI016 | Carousell Coins are publicly listed between US$2.99 and US$12.99 on the App Store. | 中 | SI012 |
| CI017 | A 3-Day Bump is publicly listed at US$2.99 on the App Store. | 中 | SI012 |
| CI018 | Buyer Protection and the Buy button expand Carousell beyond listing revenue into transaction-linked monetization. | 中 | SI013 |
| CI019 | Integrated delivery further increases the probability that Carousell earns revenue from fulfillment-adjacent services rather than only ads. | 中 | SI013 |
| CI020 | The business model has moved from pure classifieds toward a blended marketplace plus recommerce operator with both software-like and operations-heavy revenue streams. | 中 | SI001, SI002, SI004, SI013 |
| CI021 | Public disclosures do not provide gross margin, contribution margin, CAC, payback, or take-rate detail sufficient for a conventional marketplace underwriting model. | 中 | |
| CI022 | That missing unit-economics detail is a real diligence blocker because positive EBITDA can coexist with weak per-order economics or subsidized growth. | 中 | SI001, SI004 |
| CI023 | Carousell's recommerce push likely raises gross merchandise control and take-rate opportunity relative to classifieds advertising alone. | 中 | SI001, SI002, SI003 |
| CI024 | The same recommerce push also likely adds working-capital, inventory, inspection, and store-level operating complexity. | 中 | SI003, SI020, SI021 |
| CI025 | Cloudflare said Carousell handled sale-day traffic spikes of over 3x normal traffic, implying infrastructure cost and reliability matter directly to monetization. | 中 | SI020 |
| CI026 | Google Cloud said listing time fell below two minutes and items listed rose 20%, implying product efficiency can support seller supply growth without equivalent labor growth. | 中 | SI021 |
| CI027 | Google Cloud also said translation automation covers 90% of translation needs and eliminates subscription costs, indicating some operating leverage in support functions. | 中 | SI021 |
| CI028 | Carousell cut related on-call burden to zero and compute-cluster cost by 10% after migrating workloads, showing infrastructure tuning can improve margins. | 中 | SI021 |
| CI029 | Management commentary in 2023 and 2025 consistently points to disciplined spending rather than hyper-growth-at-all-costs. | 中 | SI004, SI005 |
| CI030 | The December 2024 layoff of roughly 7% of staff is consistent with continued efficiency pressure on the model. | 中 | SI004 |
| CI031 | The December 2022 layoff of 110 staff, about 10% of workforce, further shows earlier overinvestment and subsequent retrenchment. | 中 | SI005 |
| CI032 | The capital stack appears equity-funded rather than debt-heavy in reviewed public sources, but public disclosure is insufficient to rule out working-capital facilities or local credit arrangements. | 中 | SI006, SI007 |
| CI033 | Public sources do not disclose FY25 net income, EBITDA dollars, or ending cash balance, limiting confidence in runway analysis. | 中 | |
| CI034 | Because FY24 cash was US$50.5 million and management described the company as well capitalised, insolvency risk appears lower than during prior hyper-growth years but remains impossible to quantify precisely from public evidence. | 中 | SI004, SI005 |
| CI035 | The last visible valuation mark of US$1.1 billion now sits several years behind the FY25 EBITDA milestone, so the next price should depend more on proof of sustained profitability than on unicorn heritage. | 中 | SI001, SI006, SI019 |
| CI036 | Public pricing visibility is unusually thin for a company of Carousell's scale: app-store IAPs are visible, but merchant pricing, take rates, and category-level commissions mostly are not. | 中 | SI012, SI013 |
| CI037 | Compared with public marketplace comps, Carousell remains information-sparse on the exact metrics an investor would use for cohort and margin underwriting. | 中 | SI022, SI023, SI025 |
| CI038 | The financial story is improving, but the current public package supports “progress toward quality” more clearly than it supports precise intrinsic valuation. | 中 | SI001, SI004, SI005, SI006 |
| CE001 | Carousell is a multi-category recommerce and classifieds platform rather than a single-vertical marketplace. | 高 | SE001, SE016, SE017 |
| CE002 | The core user workflow still centers on fast listing, chat-based negotiation, and local commerce, but the company now layers payments, delivery, inspection, and direct sell-to-platform paths on top. | 中 | SE015, SE016, SE017 |
| CE003 | Carousell's visible product modules include marketplace listings, CarouBiz seller tools, Buyer Protection, Official Delivery, Carousell Certified, and Sell to Carousell. | 高 | SE007, SE015, SE016, SE017 |
| CE004 | Carousell supports direct sell-to-platform workflows for clothing, phones, luxury bags, and cars in selected markets. | 中 | SE016, SE017 |
| CE005 | Certified listings are publicly described for cars, mobiles, luxury bags, and in some sources luxury watches. | 高 | SE010, SE015, SE017 |
| CE006 | The group's product surface extends beyond peer listings into acquired vertical brands such as LuxLexicon and previously REFASH and Laku6. | 中 | SE018, SE025 |
| CE007 | LuxLexicon remained a separate brand, team, and retail space after acquisition, implying a multi-brand operating model rather than immediate absorption. | 中 | SE018 |
| CE008 | Carousell's luxury workflow includes both Certified Luxury and Sell to Carousell Luxury paths. | 中 | SE018 |
| CE009 | On the consumer side, Carousell emphasizes AI-assisted listing creation, price comparison, reviews, safe checkout, and trusted home services in one app. | 高 | SE016, SE017 |
| CE010 | On the seller side, CarouBiz is explicitly framed as a subscription product with additional business features across web and mobile. | 中 | SE007, SE016 |
| CE011 | Google Cloud says Carousell uses Gemini, Vertex AI, BigQuery, Cloud Storage, Compute Engine, Google Kubernetes Engine, and Google Workspace. | 中 | SE003 |
| CE012 | The AI-driven List with AI feature can auto-populate title, description, category, price, and condition from photos or short video. | 高 | SE003, SE016 |
| CE013 | Google Cloud says median listing time fell below two minutes and items listed rose 20% after the AI-led listing flow. | 高 | SE003, SE016 |
| CE014 | Carousell now runs bulk chat-risk inference using BigQuery plus Vertex AI rather than a more limited custom model. | 高 | SE003, SE013 |
| CE015 | Before the Google Cloud upgrade, the prior chat-risk model covered only 10-20% of conversations and was difficult to scale cost-effectively. | 中 | SE003 |
| CE016 | Google Cloud says translation automation now covers 90% of translation needs and replaced paid subscription services. | 中 | SE003 |
| CE017 | The same case study says workload migration to C3D instances cut cluster cost by 10% and reduced related on-call burden to zero. | 中 | SE003 |
| CE018 | Cloudflare says Carousell serves around 1 PB of images per month. | 中 | SE004 |
| CE019 | Cloudflare says flash-sale peaks can attract more than 3x normal traffic. | 中 | SE004 |
| CE020 | Cloudflare says Carousell is fully cached on Cloudflare and uses its CDN, WAF, SSL, and DNS stack. | 中 | SE004 |
| CE021 | Cloudflare reported roughly a 60% image-bandwidth cache hit ratio and around 50 Page Rules in use. | 中 | SE004 |
| CE022 | Public GitHub surfaces show Carousell engineers maintaining open-source or public-facing repositories across Go, Swift, Kotlin, JavaScript, TypeScript, Python, and shell tooling. | 高 | SE005, SE006 |
| CE023 | The public Orion microservices framework suggests an internal service-oriented engineering culture around grpc/protobuf. | 中 | SE005, SE006 |
| CE024 | Carousell maintains a dedicated safety-and-security help center and multiple safety policy pages as part of the product experience, not just corporate messaging. | 中 | SE009, SE010, SE011 |
| CE025 | Carousell's trust stack includes message blocking for links, email addresses, and QR codes, plus real-time warnings and dispute banners. | 高 | SE010, SE011, SE013 |
| CE026 | Carousell uses mobile-number verification for chat and listings and layered 2FA or SMS reverification in suspicious-login scenarios. | 高 | SE011, SE010, SE014 |
| CE027 | MHA said Carousell committed to add more secure login options such as passkeys and biometric logins by 31 January 2026. | 高 | SE014, SE022, SE023 |
| CE028 | Carousell also committed to suspend relinquished accounts and blacklist Singpass credentials tied to suspended scam accounts. | 高 | SE014, SE022 |
| CE029 | The MHA marketplace page shows that Carousell treats user badges, verified sellers, secure payments, and platform-record retention as productized trust features. | 中 | SE015 |
| CE030 | Carousell's sanctions page shows that compliance controls include geo-blocking, listing screening, and unsupported-country enforcement, indicating back-office risk tooling in addition to front-end trust tools. | 中 | SE012 |
| CE031 | CNBC and the FY25 release reinforce that AI is now a central product and trust roadmap, not a side experiment. | 中 | SE002, SE019 |
| CE032 | Google Cloud says the roadmap goal is to reduce listing time to under 45 seconds and enrich listings with stylistic attributes for better discovery. | 中 | SE003 |
| CE033 | Carousell is also exploring agentic coding tools for internal engineering tasks, suggesting continued investment in engineering productivity. | 中 | SE003 |
| CE034 | The company's physical recommerce store network means the product is partly software, partly operating process, especially for high-trust categories. | 中 | SE002, SE025 |
| CE035 | Public evidence does not support strong claims about patents, proprietary hardware, or a deep algorithmic moat beyond execution, data, and trust operations. | 中 | SE003, SE004, SE005 |
| CE036 | Public evidence also does not reveal a dedicated public status page or published uptime SLA for marketplace users. | 中 | |
| CE037 | Carousell's product differentiation appears strongest in workflow compression, trust layers, and category-specific operational assets rather than in one standalone software feature. | 中 | SE003, SE015, SE018 |
| CE038 | The product stack is mature enough to show real production scale and operational refinement, but still leaves diligence gaps around reliability metrics, incident history, and exact partner economics. | 中 | SE003, SE004, SE015 |
| CU001 | Carousell Group publicly says it serves tens of millions of users across seven markets and multiple brands. | 高 | SU001, SU003, SU013 |
| CU002 | The main Carousell marketplace is available in Singapore, Hong Kong, Indonesia, Malaysia, the Philippines, and Taiwan. | 高 | SU001, SU005 |
| CU003 | The broader group footprint includes brands such as Cho Tot, Laku6, LuxLexicon, Mudah.my, OneShift, REFASH, and Revo Financial. | 中 | SU001 |
| CU004 | Carousell customer activity spans everyday resale, property, autos, home services, luxury, mobile phones, and fashion. | 中 | SU002, SU004, SU020 |
| CU005 | The platform is structurally multi-sided: casual consumers, repeat sellers, verified businesses, and specialist merchants all coexist in the same ecosystem. | 中 | SU007, SU009, SU022 |
| CU006 | CarouBiz exists specifically for business users, showing that Carousell is not limited to purely casual peer-to-peer sellers. | 中 | SU022, SU004 |
| CU007 | Specialist Awards are used to spotlight top merchants who demonstrate exceptional quality, expertise, and customer service. | 高 | SU007, SU024, SU009 |
| CU008 | Verified Business and other user badges help buyers distinguish professional or trusted sellers from ordinary or new users. | 中 | SU009 |
| CU009 | Buyer and seller reviews are an embedded part of the product and can be left after successful transactions, including Buy-button orders. | 高 | SU006, SU021, SU004 |
| CU010 | Carousell reviews rate sellers on communication, coordination, and description accuracy, and buyers on payment coordination. | 中 | SU006 |
| CU011 | Users have 30 days to leave a review, and 4/5-star reviews can include profile-visible compliments. | 中 | SU006 |
| CU012 | Compliments such as “Goes the extra mile,” “Amazing chat,” “Fast and decisive,” and “Easygoing buyer” turn reviews into portable trust signals. | 中 | SU006 |
| CU013 | App-store surfaces describe Carousell as a place where millions of people declutter, earn money, and find deals or services. | 高 | SU004, SU005 |
| CU014 | The iOS listing shows a 4.7 out of 5 rating from 3.1K ratings. | 中 | SU004 |
| CU015 | JustUseApp says its NLP model reviewed 10,339 user reviews and still classifies Carousell as generally safe but to use with caution. | 低 | SU008 |
| CU016 | JustUseApp excerpts include positive use cases such as buying a dream car, selling a car quickly, and a fuss-free piano sale. | 低 | SU008 |
| CU017 | Carousell's own customer-proof signals therefore show high app-store ratings but also continued trust friction visible in external review aggregators. | 中 | SU004, SU008 |
| CU018 | Over 93% of active users across Carousell markets did not encounter a scammer in 2025. | 高 | SU010, SU012 |
| CU019 | 99.96% of transactions were scam-free in the 2025 regional report. | 高 | SU010, SU012 |
| CU020 | Carousell suspended over 451,000 suspicious accounts and reviewed over 520,000 community reports in 2025. | 高 | SU010, SU012 |
| CU021 | These trust statistics matter directly to customer durability because unsafe marketplaces destroy repeat usage faster than almost any other defect. | 中 | SU010, SU017, SU018 |
| CU022 | Carousell's secure checkout, official delivery, and verified-seller flows are attempts to convert fragile peer-to-peer trust into repeatable transaction behavior. | 中 | SU009, SU010, SU011 |
| CU023 | Carousell has moved part of its customer experience offline through 29 physical stores concentrated in Singapore, Hong Kong, Malaysia, and Indonesia. | 高 | SU003, SU014 |
| CU024 | DealStreetAsia said these stores and official trade-ins delivered more than 20% GMV uplift in FY25. | 中 | SU014 |
| CU025 | The store strategy likely matters most for customers buying or selling phones, luxury items, and other trust-sensitive categories. | 中 | SU014, SU023 |
| CU026 | Google Cloud says AI-driven listing reduced listing time below two minutes and increased items listed by 20%, a useful adoption proxy on the seller side. | 中 | SU016 |
| CU027 | That proxy suggests supply growth can come not only from more users but from lower friction for existing users. | 中 | SU016 |
| CU028 | Luxury and specialist merchant programs show a second adoption path: deeper engagement from higher-intent or professional sellers. | 中 | SU007, SU023, SU024 |
| CU029 | Carousell therefore serves at least four economically distinct customer segments: casual consumers, trust-sensitive buyers, specialist merchants, and business sellers. | 中 | SU004, SU007, SU009, SU022 |
| CU030 | No reviewed public source discloses GRR, NRR, annual churn, repeat purchase rate, or cohort retention by category. | 中 | |
| CU031 | That means app ratings and trust stats are useful but cannot substitute for real cohort durability analysis. | 中 | SU004, SU008, SU010 |
| CU032 | Customer concentration risk is structurally lower than in enterprise software because Carousell is consumer and small-merchant multi-sided, not dependent on a few named accounts. | 中 | SU001, SU022 |
| CU033 | However, channel and geography concentration still matter because the main marketplace footprint is limited to a handful of Asian markets. | 中 | SU001, SU005 |
| CU034 | Trust and regulatory requirements in Singapore can disproportionately affect visible customer experience because Singpass-linked safeguards and MHA scrutiny are most explicit there. | 中 | SU009, SU017, SU025 |
| CU035 | Carousell's best land-and-expand story is not enterprise seat expansion but moving users from casual listing into protected checkout, certified categories, and direct-sell flows. | 中 | SU009, SU014, SU023 |
| CU036 | The company's customer proof is strongest for broad consumer activity and trust workflows, but much weaker for retention and category-specific repeat behavior. | 中 | SU001, SU006, SU010, SU010 |
| CU037 | External review aggregation is mixed enough that customer-service and dispute-resolution quality should be a live diligence topic despite strong top-line usage. | 中 | SU008, SU018, SU019 |
| CU038 | Overall, the public evidence supports large-scale adoption and a maturing trust-led customer journey, but not a high-confidence view on retention economics. | 中 | SU001, SU010, SU014, SU014 |
| CR001 | Carousell's most visible regulatory risk is scam-related intervention under Singapore's Online Criminal Harms Act regime. | 高 | SR008, SR010, SR011 |
| CR002 | MHA said e-commerce scams on Carousell increased to 513 cases in April-June 2025 from 470 a year earlier. | 中 | SR008 |
| CR003 | MHA also said total reported scams on Carousell fell to 720 from 1,125 over the same period, showing mixed risk signals rather than a clean win. | 中 | SR008 |
| CR004 | Carousell committed to implement more secure login options such as passkeys and biometrics by 31 January 2026. | 高 | SR008, SR010, SR011 |
| CR005 | Carousell also committed to suspend relinquished accounts and blacklist Singpass credentials used to verify suspended scam accounts. | 中 | SR008 |
| CR006 | Even Singpass-verified accounts were reportedly abused by scammers, which weakens the comfort investors might take from verification alone. | 中 | SR008, SR011 |
| CR007 | The 2025 transaction safety rankings still placed Carousell near the bottom of major marketplaces in Singapore. | 中 | SR014 |
| CR008 | The same ranking article said Carousell accounted for 17.1% of reported online-platform scam cases in 2024. | 中 | SR014 |
| CR009 | Carousell's own report claims 99.96% scam-free transactions and over 93% of active users not encountering a scammer in 2025. | 高 | SR005, SR006, SR012 |
| CR010 | These trust metrics are directionally positive but do not eliminate regulatory escalation risk because MHA can still require additional controls. | 中 | SR008, SR009, SR012 |
| CR011 | Carousell collects a wide range of sensitive data including payment details, identity information, chat content, location, video/audio data, and dispute evidence. | 高 | SR001, SR021 |
| CR012 | That data surface increases privacy, security, and compliance exposure even if no public breach is disclosed in reviewed sources. | 中 | SR001, SR019 |
| CR013 | The privacy policy explicitly says Carousell may use AI in customer-service interactions and conversation analysis, which heightens model-governance and data-use scrutiny. | 中 | SR001 |
| CR014 | Trade compliance risk is also real because Carousell blocks unsupported countries and screens transactions for sanctions and anti-money-laundering concerns. | 中 | SR002, SR031 |
| CR015 | Because Carousell operates across multiple jurisdictions, policy divergence on identity verification, content, payments, and shipments can raise compliance cost. | 中 | SR001, SR002, SR008 |
| CR016 | Operationally, Carousell remains exposed to phishing, off-platform fraud, deposit scams, digital-goods scams, and counterfeit risk. | 高 | SR005, SR007, SR012, SR032 |
| CR017 | Carousell's mitigation stack includes link, email, and QR-code blocking, dispute banners, moderators, buyer protection, and CROP counterfeit takedowns. | 高 | SR006, SR007, SR009 |
| CR018 | Those mitigations reduce risk but also add operating cost, user friction, and potential false positives. | 中 | SR006, SR019 |
| CR019 | CarouBiz terms disclaim uninterrupted service and note downtime may occur from forces beyond Carousell's control, leaving some service-level risk with users. | 中 | SR003 |
| CR020 | No public source in the reviewed set provides a user-facing uptime SLA or incident history, which leaves reliability risk under-documented. | 中 | |
| CR021 | Cloud dependence is material: Google Cloud underpins AI, data, compute, and internal workflows, while Cloudflare underpins CDN, WAF, DNS, and SSL. | 高 | SR019, SR020 |
| CR022 | A major failure, cost shock, or policy issue with either cloud partner would likely flow through to availability, trust tooling, and operating margin. | 中 | SR019, SR020 |
| CR023 | Carousell also depends on app stores and external payment or delivery rails to complete many customer journeys. | 中 | SR021, SR022, SR009 |
| CR024 | Because Buyer Protection and some trusted flows are only available in selected markets and categories, customer risk varies materially by geography. | 中 | SR009, SR022 |
| CR025 | The business still carries model risk because FY25 EBITDA positivity does not mean net profitability, cash durability, or robust unit economics are proven publicly. | 中 | SR015, SR016, SR017 |
| CR026 | Carousell was still loss-making in FY24 and had recently laid off 76 employees in late 2024. | 中 | SR017 |
| CR027 | Earlier, Carousell laid off 110 staff in December 2022 after over-investment, showing that execution corrections can be severe. | 中 | SR018 |
| CR028 | Sector-wide retrenchments at Lazada, Shopee, and other platforms show that Carousell operates in an environment where profitability pressure is persistent, not idiosyncratic. | 中 | SR029 |
| CR029 | Competition risk is intense because consumer attention and seller budgets are increasingly captured by a small set of scaled platforms. | 中 | SR024, SR025, SR026 |
| CR030 | BusinessMirror cited a report saying Shopee, TikTok Shop, and Lazada control 98.8% of Southeast Asia's online retail market. | 中 | SR025 |
| CR031 | Cube said Shopee held 59% of combined platform GMV in FY2025, TikTok Shop 31%, and Lazada 9%. | 中 | SR026 |
| CR032 | TikTok Shop's structurally faster growth and live-commerce model are a meaningful threat to Carousell in discovery-led categories. | 中 | SR024, SR026, SR027 |
| CR033 | Facebook Marketplace remains a major low-friction substitute in local commerce while also ranking poorly on safety, showing that user demand can persist even on riskier platforms. | 中 | SR014, SR028 |
| CR034 | Operational intensity rises with stores, official trade-ins, inspections, and authenticated categories, which means customer trust gains may come with heavier execution burden. | 中 | SR016, SR020 |
| CR035 | Because stores and trade-ins are central to the recommerce model, failure in store economics or category-level quality control could impair both growth and margin. | 中 | SR016, SR017 |
| CR036 | Review and sentiment systems help trust, but they do not prevent poor service experiences, and external review aggregators remain mixed. | 中 | SR023, SR030 |
| CR037 | Key-person and management-execution risk remains material because the founder-CEO is still the public face of strategic and financing decisions. | 中 | SR015, SR018 |
| CR038 | Acquisition-integration risk remains present because Carousell is operating a portfolio including Laku6, LuxLexicon, REFASH, and other category brands. | 中 | SR015, SR016 |
| CR039 | Trust and safety teams may face rising workload even as AI handles more screening, because regulators still expect accountable human judgment. | 中 | SR008, SR029 |
| CR040 | The absence of full public disclosure on dispute rates, fraud loss rates, and insurance or reserve policies is a continuing diligence blocker. | 中 | |
| CR041 | Carousell's top risks are therefore not theoretical edge cases but already visible operating tensions between trust, growth, and profitability. | 中 | SR008, SR016, SR017, SR025 |
| CR042 | The thesis-break condition is not one bad quarter alone, but evidence that trust controls, transaction growth, and economics cannot improve simultaneously. | 中 | SR008, SR015, SR016, SR025 |
| CV001 | Carousell's last clearly public financing valuation remains the US$1.1 billion unicorn round announced in September 2021. | 高 | SV007, SV008 |
| CV002 | At FY25 revenue of US$141 million, that historical valuation implies roughly 7.8x revenue. | 高 | SV002, SV003, SV008 |
| CV003 | Carousell grew FY25 revenue 18% year on year and reached positive EBITDA, materially improving the quality of the old 2021 valuation anchor. | 高 | SV002, SV003, SV004 |
| CV004 | FY24 revenue was US$119.3 million and net loss was US$33.2 million, so the business entered FY25 from a still loss-making base. | 中 | SV006 |
| CV005 | Carousell therefore looks re-rated fundamentally even without a new public funding round, because scale and profitability are both better than they were in 2021. | 中 | SV002, SV003, SV006, SV008 |
| CV006 | Total capital raised is publicly tracked in roughly the mid-US$300 million range, which lowers near-term financing pressure but does not by itself validate valuation quality. | 中 | SV011, SV012, SV013 |
| CV007 | Recommerce accounted for 45% of FY25 revenue, indicating a more transaction-led and potentially more defensible mix than a pure classifieds model. | 中 | SV004, SV015 |
| CV008 | Carousell's 29 stores and multi-brand portfolio show a deeper operating model than a thin-listing marketplace, but also create more execution intensity. | 中 | SV015, SV003 |
| CV009 | Vinted is the strongest premium private comparable because it combines scale, profitability, category adjacency, and C2C behavior. | 中 | SV016, SV017, SV018 |
| CV010 | Vinted reported €1.1 billion revenue, €151 million adjusted EBITDA, and €62 million net profit for 2025. | 高 | SV016, SV017 |
| CV011 | Vinted's April 2026 secondary transaction valued it at €8 billion, implying roughly 7.3x revenue. | 高 | SV016, SV017, SV019 |
| CV012 | Back Market is another useful private comp because it monetizes higher-trust recommerce, refurbished supply, and service add-ons. | 中 | SV020 |
| CV013 | Sacra estimates Back Market generated about US$495 million revenue in 2025 at a US$5.7 billion valuation, implying roughly 11.5x revenue. | 中 | SV020 |
| CV014 | Back Market's higher implied multiple suggests investors pay up for trusted supply, service layers, and global scale in recommerce. | 中 | SV020 |
| CV015 | ThredUp is the most relevant public comp because it is a listed resale marketplace with transparent profitability and revenue disclosures. | 中 | SV021, SV022, SV025 |
| CV016 | ThredUp reported 2025 revenue of US$310.8 million and guided 2026 revenue to about US$351 million to US$356 million. | 高 | SV021, SV022, SV023 |
| CV017 | Using the user-provided July 2026 market-cap reference of roughly US$775 million, ThredUp trades around 2.2x 2026 revenue. | 中 | SV021, SV024 |
| CV018 | Carousell's implied 7.8x revenue anchor is therefore far above public resale multiples but broadly in line with premium private recommerce comps. | 高 | SV002, SV008, SV016, SV017, SV020, SV021 |
| CV019 | That spread is reasonable only if investors believe Carousell can continue compounding trust and margin without a relapse into subsidy-heavy growth. | 中 | SV002, SV003, SV026, SV027 |
| CV020 | Carousell is smaller than both Vinted and Back Market by revenue, which argues for a size and liquidity discount versus top-tier private comps. | 中 | SV002, SV016, SV020 |
| CV021 | Carousell also lacks public disclosure on GMV, take rate by category, and net profitability, which limits confidence in any premium multiple. | 中 | |
| CV022 | The company does, however, have better strategic optionality than a classifieds-only peer because it now spans stores, trade-ins, specialty verticals, and AI-led trust tooling. | 中 | SV004, SV014, SV015, SV029, SV030 |
| CV023 | Carousell's geographic concentration in Southeast Asia deserves some discount versus Vinted's broader European footprint because competitive intensity and monetization capacity are less settled. | 中 | SV001, SV016, SV026, SV028 |
| CV024 | The region's market opportunity is large enough to support meaningful upside if Carousell becomes the default trusted recommerce layer rather than just another listings app. | 中 | SV001, SV028, SV029 |
| CV025 | Competition from Shopee, TikTok Shop, and Lazada compresses that upside by making traffic acquisition and seller attention structurally more expensive. | 中 | SV026, SV027 |
| CV026 | Carousell's core underwriting tension is that valuation support has improved faster than competitive uncertainty has disappeared. | 中 | SV003, SV004, SV026, SV027 |
| CV027 | A simple revenue-multiple framework places a conservative public-like band near 3x-4x FY25 revenue, or about US$423 million to US$564 million. | 中 | SV002, SV021, SV024 |
| CV028 | A hybrid band that recognizes EBITDA positivity and strategic scarcity can support roughly 5x-7x FY25 revenue, or about US$705 million to US$987 million. | 中 | SV002, SV003, SV016, SV021 |
| CV029 | A premium private-recommerce band of about 7x-9x FY25 revenue yields roughly US$987 million to US$1.27 billion. | 中 | SV002, SV016, SV017, SV020 |
| CV030 | The old US$1.1 billion headline valuation sits near the high end of the defendable range rather than obviously above it. | 中 | SV002, SV008, SV016, SV017, SV020 |
| CV031 | That means a fresh 2026 mark can be defended, but only conditionally and with careful attention to risk-adjusted quality of earnings. | 中 | SV002, SV003, SV006, SV021 |
| CV032 | The base-case fair range in this report is US$0.9 billion to US$1.15 billion, with upside to around US$1.25 billion if FY26 execution sustains. | 中 | SV002, SV003, SV004, SV016, SV020 |
| CV033 | The bear case falls toward US$0.55 billion to US$0.8 billion if growth slows, trust costs rise, or public-market comp pressure dominates. | 中 | SV006, SV021, SV026, SV027 |
| CV034 | The bull case approaches US$1.2 billion to US$1.4 billion if EBITDA durability, store productivity, and AI-enabled conversion improvements become visible. | 中 | SV002, SV004, SV014, SV029 |
| CV035 | Carousell should not be valued like Vinted today because Vinted already shows materially greater scale, profit depth, and investor liquidity. | 高 | SV016, SV017, SV018 |
| CV036 | Carousell also should not be valued as cheaply as public resale comps if investors believe the private strategic optionality and recommerce mix are real. | 中 | SV002, SV004, SV015, SV020 |
| CV037 | On balance, the most defensible stance is that Carousell is approximately at-market to slightly above-market versus a risk-adjusted 2026 private recommerce peer set. | 中 | SV002, SV003, SV016, SV020, SV021 |
| CV038 | That stance supports a conditional-pass investment view rather than a clean pass, because the company has improved enough to merit interest but not enough to ignore disclosure and competition gaps. | 中 | SV003, SV006, SV026, SV027 |
| CV039 | The decisive diligence questions are whether EBITDA is durable, whether stores and specialty verticals earn attractive returns, and whether trust improvements continue without heavy spend inflation. | 中 | SV002, SV014, SV015, SV029 |
| CV040 | If management can evidence sustained positive EBITDA, improving category mix, and clean trust metrics through FY26, the valuation debate becomes materially easier in its favor. | 中 | SV002, SV004, SV014 |
| 编号 | 出版方 | 标题 | 引文 |
|---|---|---|---|
| SO001 | Carousell Press | Our timeline and history | Founded: August 2012, Singapore; Leading Presence: Singapore, Hong Kong, Malaysia, Indonesia, the Philippines, Taiwan, Vietnam; Team Size: 800+ teammates; Users: Tens of millions. |
| SO002 | Carousell Press | Carousell Group Reaches EBITDA Profitability Milestone as Recommerce Scales, Sets AI Priorities for FY26 | Revenue grew 18% year-on-year to $141 million... Recommerce revenue grew 40% year-on-year and accounted for 45% of total revenue in FY25... By the end of FY25, the Group operated 29 physical stores across four markets. |
| SO003 | DealStreetAsia | SG unicorn Carousell turns EBITDA positive in FY25 on recommerce revenue | The company posted a revenue of $141 million for the year, up 18% from the previous year. |
| SO004 | The Straits Times | Carousell well-capitalised, with healthy narrowing of losses: CEO | It was last valued at US$1.1 billion... An IPO has been on the cards for Carousell... discussions reportedly ended amid macroeconomic uncertainty and valuation concerns. |
| SO005 | Online Marketplaces | Carousell FY24: Losses Narrow To $33.2 Million As Revenues Edge Upwards | Revenues increase to $119.3 million... Net losses of $33.2 million were down 13% YoY... The company made 76 employees redundant in a round of layoffs in December 2024. |
| SO006 | Tracxn | Carousell company profile | Carousell has raised $376M in funding... with a current valuation of $1.1B. |
| SO007 | Tracxn | Carousell funding rounds and investors | Carousell has raised a total funding of $376M over 9 rounds... latest funding round was a Series E round on Sep 04, 2023 for $2.8M. |
| SO008 | Dealroom | Carousell — Unicorn company profile | Telenor Group 26.2%... 16 countries with team presence... 1,239 employees mapped person by person. |
| SO009 | TNGlobal | Singapore's recommerce platform Carousell reaches EBITDA profitability as revenue hits $141M | AI is already delivering measurable impact, enabling sellers to create listings in as little as three seconds and helping ensure more than 99 percent of transactions... are completed without a scam incident. |
| SO010 | Google Cloud | Carousell case study | Boosts items listed per seller by 20%... Reduces median listing time to less than two minutes... Automates 90% of translation needs. |
| SO011 | Cloudflare | Carousell secures and scales their ecommerce marketplace with Cloudflare | Carousell chief architect Harshad Rotithor estimates that more than a quarter of Singapore’s entire population uses the platform, which boasts more than 250 million listings to date. |
| SO012 | GitHub | Carousell organization profile | Orion is a small lightweight framework written around grpc/protobuf with the aim to shorten time to build microservices at Carousell. |
| SO013 | GitHub | Carousell repositories | Repositories list shows internal engineering libraries and active updates through July 2026. |
| SO014 | Carousell Blog | Carousell Blog - Get tips and hear stories from your Community | Carousell publishes category content spanning HDB renovation, used cars, and property-related guides. |
| SO015 | Apple App Store | Carousell: Snap-Sell, Chat-Buy App | Carousell is the leading community marketplace in Singapore, Hong Kong, Taiwan, Malaysia, the Philippines, and Indonesia. |
| SO016 | Google Play | Carousell: Sell and Buy | Millions of people use Carousell to declutter and sell to earn money, or find great deals and trusted services. |
| SO017 | Carousell Help Centre | Privacy Policy | We may collect account credentials, identity information, location information, communications, and information required for payment, shipping, and buyer protection services. |
| SO018 | Carousell Help Centre | CarouBiz - Terms & Conditions | CarouBiz subscriptions govern access to additional seller features on Carousell's website, mobile applications and services. |
| SO019 | The Company Check | Carousell — Company Profile | Carousell has raised $375.6 million across 9 funding rounds... Annual revenue $115.94M as on Dec 31, 2023. |
| SO020 | DMR | 15 Interesting Carousell Facts and Statistics | Carousell’s press fact sheet describes the platform as having “tens of millions” of monthly active users and a team size of 1,000+ teammates. |
| SO021 | Carousell Press | Carousell Press | Press site highlights newsroom, green impact, and about our group materials. |
| SO022 | Startup Intros | Carousell: Funding, Team & Investors | Carousell: Online mobile marketplace for recommerce, facilitating buying and selling used goods for individuals in Southeast Asia and beyond. |
| SO023 | ACRA BizFile | Bizfile | Bizfile |
| SO024 | Carousell Press | Carousell Group raises US$100 million from STIC Investments to accelerate Classifieds 4.0 in Greater Southeast Asia | This round of funding, led by STIC Investments... will bring Carousell’s valuation to US$1.1 billion. |
| SO025 | EDBI | Carousell Group becomes Southeast Asia's latest unicorn with a US$100 million funding round to fuel expansion and drive sustainable commerce | This round of funding, led by STIC Investments, marks a significant milestone that will bring Carousell’s valuation to US$1.1 billion. |
| SM001 | Carousell Press | Our timeline and history | Marketplace categories: Cars, Luxury, Mobile Phones, Fashion, Home & Living, Property, Services. |
| SM002 | Carousell Press | Carousell Group Reaches EBITDA Profitability Milestone as Recommerce Scales, Sets AI Priorities for FY26 | Driven by growing consumer demand for more convenient and trusted secondhand transactions, the Company has expanded beyond traditional classifieds. |
| SM003 | DealStreetAsia | SG unicorn Carousell turns EBITDA positive in FY25 on recommerce revenue | The numbers point to Carousell’s ongoing attempt to move beyond its roots as a classifieds marketplace into a more transaction-led secondhand commerce platform. |
| SM004 | TNGlobal | Singapore's recommerce platform Carousell reaches EBITDA profitability as revenue hits $141M | Founded in August 2012 in Singapore, Carousell Group operates across seven markets... serving tens of millions of users. |
| SM005 | TMO Group | SEA eCommerce: Sales Data by Country & Industry (+ FREE Reports) | In June 2024, total sales revenue of Shopee and Lazada in Indonesia reached approximately $1.16 billion... Vietnam reached approximately $1.49 billion. |
| SM006 | Statista | Topic: E-commerce in Southeast Asia | E-commerce market value 181bn USD... high shares of internet users in selected Southeast Asian markets reported buying products on social media in 2025. |
| SM007 | Sellercraft | Southeast Asia’s Online Retail Outlook (2025–2026) | E-commerce GMV expected to reach US$230 billion by 2026, nearly doubling since 2021. |
| SM008 | Hashmeta | E-commerce Statistics Southeast Asia: Complete Market Size Analysis and Growth Insights | The region’s digital economy reached $218 billion in gross merchandise value in 2023, with e-commerce accounting for approximately $139 billion. |
| SM009 | AFFiNCO | Southeast Asia E-Commerce Statistics 2026 Full Data Report | Southeast Asia e-commerce hit $234 billion in gross merchandise value by 2026, growing 18% year over year. |
| SM010 | UnivDatos | Southeast Asia E-Commerce Market Size, Trends & Forecast 2033 | The Southeast Asia E-Commerce market was valued at USD 201.92 billion in 2024 and is expected to grow at a strong CAGR of around 21.13% during 2025-2033. |
| SM011 | DMR | 15 Interesting Carousell Facts and Statistics | Carousell is a mobile-first recommerce marketplace that helps people buy and sell items across categories like electronics, fashion, home goods, vehicles, and more. |
| SM012 | Apple App Store | Carousell: Snap-Sell, Chat-Buy App | Leading community marketplace in Singapore, Hong Kong, Taiwan, Malaysia, the Philippines, and Indonesia. |
| SM013 | Google Play | Carousell: Sell and Buy | Shop Certified listings and use escrow-protected Buyer Protection. |
| SM014 | Facebook Marketplace: Buy and Sell Items Locally or Shipped | Buy and Sell Items Locally or Shipped | |
| SM015 | eBay | Electronics, Cars, Fashion, Collectibles & More | eBay, the world's online marketplace. |
| SM016 | Mercari | Your Go-to Marketplace for Deals on Used & Secondhand Items | Your Go-to Marketplace for Deals on Used & Secondhand Items. |
| SM017 | OLX | OLX is part of OLX Group | Join the millions who buy and sell from each other everyday in local communities around the world. |
| SM018 | Quikr | Quikr | Quikr home page. |
| SM019 | OfferUp | OfferUp | OfferUp home page. |
| SM020 | Craigslist | craigslist | craigslist |
| SM021 | Wallapop | Wallapop | Wallapop home page. |
| SM022 | Depop | Depop | Depop home page. |
| SM023 | The Company Check | Carousell — Company Profile | Carousell operates in a competitive market with competitors including eBay, Quikr, Mapan, OLX and 58City. |
| SM024 | Tracxn | Carousell company profile | The company has 113 active competitors. |
| SM025 | Dealroom | Carousell — Unicorn company profile | eCommerce & Marketplaces · Singapore, Singapore · Founded 2012 Unicorn. |
| SP001 | Carousell Press | Carousell fact sheet | Carousell is one of the world’s largest and fastest growing marketplaces in classifieds. |
| SP002 | Carousell Press | FY25 profitability announcement | Set to transform the Company’s marketplace experience... accelerate AI at scale. |
| SP003 | DealStreetAsia | SG unicorn Carousell turns EBITDA positive in FY25 on recommerce revenue | Attempt to move beyond its roots as a classifieds marketplace into a more transaction-led secondhand commerce platform. |
| SP004 | Facebook Marketplace | Buy and Sell Items Locally or Shipped. | |
| SP005 | eBay | eBay home page | Electronics, Cars, Fashion, Collectibles & More. |
| SP006 | Mercari | Mercari marketplace | Marketplace for deals on used & secondhand items. |
| SP007 | OLX | OLX | Millions who buy and sell from each other everyday in local communities. |
| SP008 | Quikr | Quikr | Quikr home page. |
| SP009 | OfferUp | OfferUp | OfferUp home page. |
| SP010 | Craigslist | Craigslist | craigslist |
| SP011 | Wallapop | Wallapop | Wallapop home page. |
| SP012 | Depop | Depop | Depop home page. |
| SP013 | Similarweb | carousell.com competitors | carousell.com top competitors include ebay.com, offerup.com, craigslist.org and mercari.com. |
| SP014 | Craft | Carousell competitors | Carousell’s competitors and similar companies include eBay, quikr, Mercari and Kaidee. |
| SP015 | Tracxn | Carousell competitors | The company has 113 active competitors. |
| SP016 | Dealroom | Carousell profile | eCommerce & Marketplaces · Unicorn. |
| SP017 | eBay | eBay research tools | Use Terapeak research tools to understand what sells. |
| SP018 | Apple App Store | Carousell app | Leading community marketplace... One-stop destination for buying and selling. |
| SP019 | Google Play | Carousell app | Certified listings and escrow-protected Buyer Protection. |
| SP020 | Carousell Support | Privacy Policy | Privacy Policy. |
| SP021 | Carousell Support | Safety and security category | Safety and security. |
| SP022 | MHA Singapore | Enhanced verification measures by Carousell under strengthened marketplace safety ratings framework | Carousell has committed to implement stronger user verification measures. |
| SP023 | Google Cloud | Carousell customer story | Carousell uses Google Cloud to deliver AI-powered experiences at scale. |
| SP024 | Cloudflare | Carousell case study | Carousell improved reliability and security with Cloudflare. |
| SP025 | TMO Group | SEA ecommerce monthly update | Electronics and fashion are major online categories across Southeast Asia. |
| SP026 | Kaidee | Kaidee marketplace | Kaidee home page. |
| SP027 | TikTok Shop | TikTok Shop | TikTok Shop page returned limited accessible content but confirms shop surface. |
| SI001 | Carousell Press | Carousell Group Reaches EBITDA Profitability Milestone as Recommerce Scales, Sets AI Priorities for FY26 | Carousell reported record revenue of US$141 million in FY25 and achieved positive EBITDA for the first time. |
| SI002 | DealStreetAsia | SG unicorn Carousell turns EBITDA positive in FY25 on recommerce revenue | Revenue grew 18% to US$141 million, while recommerce revenue climbed nearly 40% and accounted for 45% of total revenue. |
| SI003 | TNGlobal | Singapore's recommerce platform Carousell reaches EBITDA profitability as revenue hits $141M | The company recorded $141 million in revenue in FY25, with recommerce contributing 45 percent of total revenue. |
| SI004 | Online Marketplaces | Carousell FY24: Losses Narrow To $33.2 Million As Revenues Edge Upwards | Revenues increase to $119.3 million ... net losses of $33.2 million ... cash and cash equivalents increased to $50.5 million. |
| SI005 | The Straits Times | Carousell well-capitalised, with healthy narrowing of losses: CEO | Carousell was last valued at US$1.1 billion ... revenue grew 66.9 per cent to US$82.5 million in 2022, but losses widened to US$63.1 million. |
| SI006 | Tracxn | Carousell funding and investors | Carousell has raised a total of $376M over 9 funding rounds. |
| SI007 | The Company Check | Carousell company profile | Annual revenue $115.94M as on Dec 31, 2023; EBITDA -$35.54M; total equity funding $375.6M. |
| SI008 | Tracxn | Carousell company profile | Carousell is a unicorn in Singapore in eCommerce & marketplaces. |
| SI009 | Dealroom | Carousell company profile | Carousell is categorized as an eCommerce & marketplaces unicorn. |
| SI010 | Startup Intros | Carousell overview | Carousell company overview page. |
| SI011 | Carousell Press | Carousell fact sheet | Carousell Group is the leading multi-category platform for secondhand in Greater Southeast Asia. |
| SI012 | Apple App Store | Carousell: Snap-Sell, Chat-Buy App | CarouBiz Lite $14.99; Carousell Coins $2.99-$12.99; Get a 3-Day Bump $2.99. |
| SI013 | Google Play | Carousell: Sell and Buy | Upgrade to CarouBiz to run your business on Carousell and tap the Buy button for secure checkout with Buyer Protection. |
| SI014 | Carousell Press | Carousell raises US$100 million from STIC-led consortium | Carousell joins the ranks of unicorns after raising US$100 million. |
| SI015 | EDBI | Carousell becomes Southeast Asia's latest unicorn | US$100 million investment led by STIC Investments. |
| SI016 | PR Newswire | Carousell raises US$80 million in new funding led by Naver | Carousell Group has raised US$80 million in new funding led by Naver. |
| SI017 | Ventures Rakuten | Rakuten portfolio | Rakuten Ventures lists Carousell in its portfolio. |
| SI018 | Rakuten Ventures | Carousell portfolio listing | Rakuten Ventures lists Carousell in its portfolio. |
| SI019 | The Straits Times | Carousell turns unicorn after raising US$100m | Carousell turns unicorn after raising US$100m in STIC-led funding round. |
| SI020 | Cloudflare | Carousell case study | Carousell handles 3x traffic during flash sales and sees around 60% cache hit ratio for image bandwidth. |
| SI021 | Google Cloud | Carousell case study | Carousell reduced median listing time to less than two minutes and increased items listed by 20%. |
| SI022 | SEC | eBay EDGAR entity landing page | SEC EDGAR page for eBay filings. |
| SI023 | eBay Investor Relations | Financial summary | Financial summary page for eBay investor relations. |
| SI024 | Caplight | Vinted valuation and funding | Vinted operates a secondhand marketplace with buyer protection fee and shipping/logistics features. |
| SI025 | Yahoo Finance | eBay key statistics | Yahoo Finance key statistics page for eBay. |
| SE001 | Carousell Press | Carousell fact sheet | Leading multi-category platform for secondhand in Greater Southeast Asia. |
| SE002 | Carousell Press | FY25 profitability and FY26 AI priorities | The company set AI priorities for FY26 while expanding transaction-led recommerce. |
| SE003 | Google Cloud | Carousell case study | Gemini reduces listing time to less than two minutes and boosts items listed by 20%. |
| SE004 | Cloudflare | Carousell case study | Carousell serves around 1 PB of images per month and handles flash-sale traffic over 3x normal volume. |
| SE005 | GitHub | Carousell organization overview | Popular repositories include Orion, iOS tooling, and developer setup scripts. |
| SE006 | GitHub | Carousell repositories | Carousell repositories show recent updates across Go, TypeScript, Swift, and infrastructure tools. |
| SE007 | Carousell Help Centre | CarouBiz Terms & Conditions | CarouBiz subscriptions provide additional features on website and mobile applications with monthly or yearly renewal. |
| SE008 | Carousell Help Centre | Privacy Policy | Privacy policy governs the collection and use of user information across services. |
| SE009 | Carousell Help Centre | Safety and security category | Carousell maintains a dedicated Safety and security help-centre category. |
| SE010 | Carousell Help Centre | How Carousell Protects You From Scams | Carousell uses AI, data science, and machine learning to detect suspicious activity 24/7. |
| SE011 | Carousell Help Centre | How Carousell has been working to keep our users safe | Multi-layered trust and safety includes ML detection, dispute banners, 2FA, mobile verification, and the CROP anti-counterfeit programme. |
| SE012 | Carousell Help Centre | Our Commitment to Trade Compliance and Sanctions | Carousell blocks unsupported regions and screens listings and transactions for compliance risks. |
| SE013 | Carousell Press | Regional Scam Trends Report 2025 | Carousell continually enhances proprietary AI, machine learning, and data science tools to detect scams and suspicious activity at scale. |
| SE014 | MHA Singapore | Enhanced Verification Measures by Carousell Under the E-Commerce Code of the Online Criminal Harms Act | Carousell committed to implement passkey or biometric logins, suspended-account blacklists, and other safeguards by 31 January 2026. |
| SE015 | MHA Singapore | Carousell marketplace transaction safety ratings page | Sell to Carousell, Carousell Certified, Buy button, official delivery, and user badges are described on the MHA page. |
| SE016 | Apple App Store | Carousell: Snap-Sell, Chat-Buy App | Take a photo and AI fills in title, description and price; sell directly to Carousell; safe checkout with Buyer Protection. |
| SE017 | Google Play | Carousell: Sell and Buy | Certified listings and official delivery are available in selected markets, while Buyer Protection supports secure checkout. |
| SE018 | Vulcan Post | Carousell doubles down on luxury segment with acquisition of LuxLexicon | Carousell launched Certified Luxury and Sell to Carousell Luxury in 2023 and later acquired LuxLexicon. |
| SE019 | CNBC | Carousell CEO: AI improving trust on our platform | Carousell CEO discussed AI improving trust on the platform. |
| SE020 | Carousell Blog | Carousell blog | The blog contains category-specific content for cars, property, and home renovation use cases. |
| SE021 | Carousell Press | Carousell home / newsroom | Carousell maintains a dedicated press site for product and business updates. |
| SE022 | Channel NewsAsia | Carousell scams and enhanced verification | Carousell will introduce stronger authentication and blacklisting safeguards after scammers abused verified accounts. |
| SE023 | The Straits Times | Carousell to step up moves to block account takeovers | Stronger login options such as passkeys and biometrics were among the measures discussed. |
| SE024 | The Independent Singapore | Fraud cases on Carousell have fallen by 36% year on year | Fraud cases on Carousell were reported down year on year as safeguards tightened. |
| SE025 | DealStreetAsia | Carousell FY25 article | Carousell operates 29 physical stores across Singapore, Hong Kong, Malaysia, and Indonesia. |
| SU001 | Carousell Press | About Carousell Group | Carousell Group serves tens of millions of users across seven markets and multiple brands. |
| SU002 | Carousell Press | Carousell fact sheet | Carousell is a multi-category platform spanning cars, property, luxury, fashion, mobile phones, and services. |
| SU003 | Carousell Press | FY25 profitability announcement | Carousell Group serves tens of millions of users and operates 29 physical stores. |
| SU004 | Apple App Store | Carousell: Snap-Sell, Chat-Buy App | Millions of people use Carousell to declutter and sell to earn money or find great deals and trusted services. |
| SU005 | Google Play | Carousell: Sell and Buy | Leading community marketplace in Singapore, Hong Kong, Taiwan, Malaysia, the Philippines, and Indonesia. |
| SU006 | Carousell Help Centre | How do I leave a review for a buyer or seller? | Both buyers and sellers can leave reviews, rate criteria, and compliments after successful transactions. |
| SU007 | Carousell College | Carousell Specialist Award | The Specialist Awards honour top merchants who demonstrate exceptional quality, expertise, and customer service. |
| SU008 | JustUseApp | Carousell reviews (2026) | Analysis of 10,339 user reviews says Carousell appears generally safe but use with caution. |
| SU009 | MHA Singapore | Carousell marketplace transaction safety ratings page | User badges, verified business labels, buy button, and review checks are part of safer transacting on Carousell. |
| SU010 | Carousell Help Centre | How Carousell Protects You From Scams | 99.96% of transactions proceeded without scam incident and over 93% of active users did not encounter a scammer. |
| SU011 | Carousell Help Centre | How Carousell has been working to keep our users safe | Ratings and reviews help to build trust and buyers or sellers can rate each other on transaction quality. |
| SU012 | Carousell Press | Regional Scam Trends Report 2025 | Over 93% of active users across all Carousell markets did not encounter a scammer. |
| SU013 | TechNode Global | Carousell reaches EBITDA profitability as revenue hits $141M | Carousell operates across seven markets, serving tens of millions of users. |
| SU014 | DealStreetAsia | Carousell turns EBITDA positive in FY25 | The group now runs 29 physical stores and saw over 20% GMV uplift from stores and official trade-ins. |
| SU015 | Expanded Ramblings / DMR | Carousell statistics (2026) | Carousell statistics page summarizes country and user metrics, including over 30 million users and coverage across major Asian markets. |
| SU016 | Google Cloud | Carousell case study | AI-led listing flow increased items listed by 20% and cut median listing time below two minutes. |
| SU017 | Channel NewsAsia | Carousell scams and enhanced verification | Enhanced verification and safer account controls are being introduced after scam incidents on verified accounts. |
| SU018 | The Straits Times | Carousell to step up moves to block accounts including Singpass-verified ones | Carousell plans stronger defenses against misuse of verified accounts. |
| SU019 | The Independent Singapore | Fraud cases on Carousell have fallen by 36% year on year | Fraud cases on Carousell were reported down 36% year on year. |
| SU020 | Carousell Blog | Carousell blog | Carousell publishes category-specific content for cars, property, and home-related users. |
| SU021 | Carousell Help Centre | Reviews section | Carousell maintains a dedicated reviews help-centre section. |
| SU022 | Carousell Help Centre | CarouBiz Terms & Conditions | CarouBiz subscriptions are intended for business users and renew monthly or yearly. |
| SU023 | Vulcan Post | Carousell acquires LuxLexicon | Carousell is extending gains in the luxury resale market and building customer confidence through authentication. |
| SU024 | Carousell College | Specialist awards for merchants | Specialist merchants are recognized for exceptional quality, expertise, and customer service. |
| SU025 | MHA Singapore | Enhanced verification measures under OCHA | High-risk sellers and advertisers were verified against government-issued records and further safeguards were mandated. |
| SU026 | NearMe.SG | Carousell Buyer Protection Singapore 2026 guide | Buyer Protection holds payment until buyers confirm receipt and works best with tracked delivery. |
| SU027 | EbbSing | Carousell Singapore guide landing page | Minimal landing page captured during fetch; retained only as a low-value external guide reference. |
| SR001 | Carousell Help Centre | Privacy Policy | Carousell collects identity, payments, chat content, location, SingPass-linked verification data, and dispute evidence. |
| SR002 | Carousell Help Centre | Our Commitment to Trade Compliance and Sanctions | Carousell blocks unsupported regions and screens listings and transactions for sanctions and financial-crime risk. |
| SR003 | Carousell Help Centre | CarouBiz Terms & Conditions | CarouBiz services are provided as-is and may experience planned or unplanned downtime beyond Carousell’s control. |
| SR004 | Carousell Help Centre | Safety and security category | Carousell maintains a dedicated safety and security help surface. |
| SR005 | Carousell Press | Regional Scam Trends Report 2025 | 99.96% of transactions were scam-free, but scammers evolved toward QR code and off-platform tactics. |
| SR006 | Carousell Help Centre | How Carousell Protects You From Scams | Over 451,000 suspicious accounts were suspended and over 520,000 reports investigated in 2025. |
| SR007 | Carousell Help Centre | How Carousell has been working to keep our users safe | Carousell uses ML detection, dispute banners, 2FA, mobile verification, and CROP anti-counterfeit partner tooling. |
| SR008 | MHA Singapore | Enhanced Verification Measures under OCHA | E-commerce scams on Carousell rose to 513 cases from April to June 2025 versus 470 a year earlier, leading to further mandated safeguards. |
| SR009 | MHA Singapore | Carousell transaction safety ratings page | Carousell requires mobile verification for listing or chat and uses badges, Buy button, and official delivery as risk controls. |
| SR010 | Channel NewsAsia | Carousell scams and enhanced verification | Carousell must strengthen account safeguards after scammers abused verified accounts. |
| SR011 | The Straits Times | Carousell to step up moves to block account takeovers | Stronger safeguards are needed because scammers used even Singpass-verified accounts. |
| SR012 | The Online Citizen | Carousell reports over 99% scam-free transactions amid evolving fraud tactics in 2025 | Singapore recorded 99.97% scam-free transactions, but phishing, digital goods, and deposit scams remained active. |
| SR013 | The Independent Singapore | Fraud cases on Carousell have fallen by 36% year on year | Fraud cases on Carousell were reported down 36% year on year. |
| SR014 | The Independent Singapore | Carousell and Facebook Marketplace remain lowest-ranked in 2025 scam safety ratings | Carousell received two ticks and accounted for 17.1% of reported online-platform scam cases in 2024. |
| SR015 | Carousell Press | FY25 profitability milestone | Carousell reached positive EBITDA in FY25 while pushing AI priorities for FY26. |
| SR016 | DealStreetAsia | Carousell turns EBITDA positive in FY25 | Carousell is still transitioning from classifieds roots to transaction-led recommerce. |
| SR017 | Online Marketplaces | Carousell FY24 losses narrow to $33.2 million | Carousell remained loss-making in FY24 and laid off 76 employees in late 2024. |
| SR018 | The Straits Times | Carousell well-capitalised, with healthy narrowing of losses | Carousell revenue grew but losses widened in 2022 and 110 staff were laid off in December 2022. |
| SR019 | Google Cloud | Carousell case study | Carousell relies on Google Cloud across AI, data, compute, and internal workflows. |
| SR020 | Cloudflare | Carousell case study | Cloudflare provides CDN, WAF, DNS, SSL, and DDoS protection for Carousell during 3x traffic spikes. |
| SR021 | Apple App Store | Carousell iOS app listing | The app collects purchases, financial info, contact info, user content, diagnostics, and tracking data. |
| SR022 | Google Play | Carousell Android app listing | Buyer Protection and certified listings are available only in selected categories and markets. |
| SR023 | JustUseApp | Carousell reviews 2026 | Carousell appears generally safe, but use with caution, based on analysis of 10,339 user reviews. |
| SR024 | Digital in Asia | SEA e-commerce 2026 hub | Southeast Asia e-commerce is on track for US$215 billion by 2026 and social commerce is becoming a quarter of the market. |
| SR025 | BusinessMirror | Three platforms control Southeast Asia online retail market | Shopee, TikTok Shop and Lazada collectively control 98.8% of the region’s online retail market. |
| SR026 | Cube | Shopee, Lazada, and TikTok Shop in Southeast Asia: What the Data Shows in 2026 | Shopee held 59% of combined GMV in FY2025, TikTok Shop 31%, and Lazada 9%. |
| SR027 | TikTok Shop | TikTok Shop | TikTok Shop page was rate-limited, but the official commerce surface exists and is regionally important. |
| SR028 | Facebook Marketplace | Facebook Marketplace remains a major local-commerce substitute. | |
| SR029 | The Straits Times | Recent layoffs at e-commerce firms reflect sector maturation | E-commerce platforms are prioritising profitability, operational efficiency, and AI, with recent layoffs across the sector. |
| SR030 | Carousell Help Centre | How to leave a review for a buyer or seller | Reviews are reciprocal and visible, but only after a transaction flow has completed. |
| SR031 | Carousell Help Centre | Terms of Service | Carousell may suspend accounts, enforce policies, and modify the service terms across jurisdictions. |
| SR032 | Carousell Help Centre | [Singapore] List of Prohibited Content | Carousell maintains granular prohibited-content rules in Singapore, increasing moderation and enforcement burden. |
| SV001 | Carousell Press | Fact Sheet | Carousell operates across Southeast Asia, Hong Kong, and Taiwan. |
| SV002 | Carousell Press | Carousell reaches EBITDA profitability milestone | FY25 revenue reached US$141 million and the group achieved positive EBITDA. |
| SV003 | DealStreetAsia | Carousell turns EBITDA positive in FY25 | Carousell reported FY25 revenue of $141 million, up 18%, and turned EBITDA positive. |
| SV004 | TechNode Global | Singapore's recommerce platform Carousell reaches EBITDA profitability as revenue hits $141M | Recommerce accounted for 45% of revenue and the group focused on AI for FY26. |
| SV005 | The Straits Times | Carousell well-capitalised, with healthy narrowing of losses | Carousell said it was well capitalised as losses narrowed and revenue grew. |
| SV006 | Online Marketplaces | Carousell FY24 losses narrow to $33.2 million | FY24 revenue was $119.3 million and net loss was $33.2 million. |
| SV007 | Carousell Press | Carousell Group raises US$100 million, joins the ranks of unicorns | Carousell became a unicorn after a US$100 million funding round. |
| SV008 | EDBI | Carousell Group becomes Southeast Asia's latest unicorn with a US$100 million funding round | The financing valued Carousell at US$1.1 billion. |
| SV009 | Rakuten Ventures | Portfolio page | Rakuten Ventures lists Carousell in its portfolio. |
| SV010 | Carousell Help Centre | How Carousell has been working to keep our users safe | Carousell describes trust, moderation, and counterfeit controls that can support marketplace quality. |
| SV011 | Tracxn | Carousell funding profile | Tracxn tracks total funding near the mid-hundreds of millions. |
| SV012 | Dealroom | Carousell company profile | Dealroom tracks Carousell as a Southeast Asian marketplace unicorn. |
| SV013 | Startup Intros | Carousell profile | Startup Intros tracks Carousell's funding history and investors. |
| SV014 | CNBC | Carousell CEO on AI improving trust on our platform | Management highlighted AI as a driver of trust and efficiency. |
| SV015 | Carousell Press | Carousell Group | Carousell Group includes Carousell Marketplace, Carousell Media Group, REFASH, Laku6, Ox Street, LuxLexicon and 29 stores. |
| SV016 | Vinted | Financial results 2025 | Vinted reported €10.8bn GMV, €1.1bn revenue, €151m adjusted EBITDA, and €62m net profit for 2025. |
| SV017 | Ontario Teachers' | Vinted secondary share transaction at €8B equity valuation | The secondary transaction valued Vinted at €8 billion and was significantly oversubscribed. |
| SV018 | CNBC | Vinted boss sees fundamental shift in consumer patterns | CNBC described Vinted as having roughly a $9 billion valuation as secondhand demand grew. |
| SV019 | Retail Gazette | Vinted valuation soars to €8bn | Vinted's €8bn valuation was up from the €5bn level referenced in an earlier secondary sale. |
| SV020 | Sacra | Back Market revenue, valuation & funding | Sacra estimates Back Market generated $495M revenue in 2025, with a $5.7B valuation and EBITDA break-even. |
| SV021 | ThredUp IR | ThredUp Q1 2026 results | ThredUp guided to full-year 2026 revenue of about $351M-$356M and around 6.1% adjusted EBITDA margin. |
| SV022 | ThredUp IR | ThredUp full-year 2025 results | ThredUp reported 2025 revenue of $310.8M and 1.65M active buyers. |
| SV023 | ThredUp IR | ThredUp full-year 2025 results PDF | The PDF details quarterly and full-year 2025 results including revenue, active buyers and cash. |
| SV024 | Stock Analysis | ThredUp revenue 2018-2026 | Stock Analysis tracks ThredUp revenue history through 2026 estimates. |
| SV025 | SEC | ThredUp 2025 annual report XBRL viewer | SEC-hosted annual report filing for ThredUp fiscal 2025. |
| SV026 | BusinessMirror | Three platforms control SEA online retail market | Shopee, TikTok Shop and Lazada collectively control 98.8% of Southeast Asia online retail according to the cited report. |
| SV027 | Cube | Shopee, Lazada, and TikTok Shop data in 2026 | Shopee held 59% of combined GMV in FY2025, TikTok Shop 31%, and Lazada 9%. |
| SV028 | Digital in Asia | SEA e-commerce hub | Digital in Asia projects Southeast Asia e-commerce to around US$215B by 2026. |
| SV029 | Google Cloud | Carousell case study | Google Cloud described Carousell as using data and AI to streamline recommendations and operations. |
| SV030 | Cloudflare | Carousell case study PDF | Cloudflare described supporting Carousell during 3x traffic spikes with CDN, WAF, DNS, and DDoS protection. |