图像一致性 Image consistency
对比主图、详情图、买家晒图和社媒帖子，找出同款、换壳和盗图。 Compare hero images, detail photos, buyer photos, and social posts to identify same products, relabels, and stolen images.
买家助手 Buyer assistant
买到靠谱商品，不花冤枉钱。 Buy something reliable without overpaying.
输入商品名称，我们会把合理价格、安全警示、真实评论和卖家可靠性放在一起比较。 Enter a product name to compare fair price, safety warnings, real reviews, and seller reliability in one place.
核对具体型号的官方召回Check an exact model for official recalls 可选工具：有完整型号或批次时再用Optional: use this when you have the full model or batch
先比价，再下单
Compare first, buy smarter
同款识别、合理价、图货一致性和卖家风险在一个页面里看清。
Same-item matching, fair price, image consistency, and seller risk in one shopping view.
低价预警
Price alerts
买家图证据
Buyer photos
卖家可靠性
Seller trust
完全成本 × 需求匹配Fully loaded cost × needs fit
价值地图Value map
官方产品实测Official product measurements
使用成本对比Operating cost explorer
—个认证型号certified models
为什么不是单纯比价 Why this is not only price comparison
便宜但不靠谱，不算便宜。 Cheap is not cheap if it does not arrive as pictured.
AI 图片风险 AI-image risk
如果主图明显是 AI 渲染、材质不合理或细节漂移，先降权，再看买家实拍。 If the listing appears AI-rendered or physically inconsistent, downgrade it and look for real buyer photos.
履约记录 Fulfillment history
把“未按图发货”“缺件”“超时”“拒绝退货”等评价聚合成卖家可靠性分。 Aggregate complaints such as wrong item shipped, missing parts, late delivery, and refund friction into a seller reliability score.
系统应该怎么工作 How the system should work
- 抓取商品、评论图、帖子、短视频标题和价格历史。 Ingest listings, review photos, posts, video titles, and price history.
- 用图片和文本把“同款”聚成一组，不只按标题搜索。 Cluster same products by image and text, not title search alone.
- 估算合理价格区间，区分工厂价、零售价、带货价和虚高券后价。 Estimate fair price bands across factory, retail, influencer, and coupon-inflated pricing.
- 给每个卖家生成可信度解释，而不是只给一个分数。 Explain seller reliability instead of only showing a score.
AI 采购助手 AI sourcing assistant
今天想完成哪件事？ What do you want to accomplish?
选一个目标，客服会一步一步带你完成，不需要先填表。 Choose a goal. The sourcing agent will guide you step by step, with no form to complete first.
AI 询价客服AI sourcing agent
在线，正在整理询价需求Online and building your RFQ
AI 先整理需求；报价、认证与生产能力仍需人工核验。 AI structures the request; a person must still verify quotes, compliance, and capacity.
利润决策Profit decision用真实供应商、物流和履约报价回答“能不能赚钱”Answer “can this make money?” from recorded supplier, freight, and fulfillment quotes
高级成本工作表Advanced cost worksheet仅在需要核对或修正自动填入的成本时打开Open only to audit or correct costs populated from evidence
项目与供应商报价Projects and supplier quotes需要时再打开管理记录Open when you need to manage records
项目组合Deal pipeline
供应商报价Supplier quotes
记录供应商报价Record supplier quote
我的采购项目My sourcing project查看项目进度、报价和下一步See progress, quotes, and the next action
当前项目Current project
项目进度Project progress
全部采购工具All sourcing tools制造商、市场、供应网络与目录Manufacturers, markets, networks, and directory
产品到工厂Product to factory
为产品寻找制造商Find manufacturers for a product
市场到履约Market to fulfillment
跨境产品机会Cross-border product opportunities
实体、设施与时效Entities, facilities, and timing
供应网络目录Supply network directory
需求到供给Demand to supply
趋势采购机会Trend-driven sourcing opportunities
产品趋势图Product trend view
热度对比Momentum comparison
市场证据Market evidence
信号来源Signal sources
| 候选方Candidate | 国家Country | 类目Categories | 起订量MOQ | 履约方式Fulfillment | 核验层级Evidence tier | 候选匹配Candidate fit | 证据Evidence |
|---|
商品研究PRODUCT RESEARCH
研究可能卖什么Research what may sell
先看今天是否有值得进一步研究的商品；没有，就看最接近达标的候选以及还缺什么。First see whether any product is worth further research today. If none is, see the closest candidates and exactly what evidence is still missing.
真实数据LIVE COLLECTION
正在读取Reading data
抠求求实际收集了什么What Kouqiuqiu has actually collected
这里显示数据库中保留的评论、商品、社媒和历史观测，而不是虚构的热度分。Retained reviews, products, social observations, and history from the database—not a made-up popularity score.
证据健康Evidence health覆盖率、时间和阻断条件Coverage, freshness, and blocking gates+
研究线索必须通过身份、需求、评论质量、竞争、混杂因素、来源健康和前瞻测试门槛，才能成为合格候选。供应和利润研究只在此之后开始。当前排名不等于销量。A research lead must pass identity, demand, review-quality, competition, confounder, source-health, and forward-test gates before it becomes a qualified candidate. Supply and profit research begins only afterward. Current rank is not sales.
卖货第 3 步 SELLING STEP 3
怎么发货，货放哪里？ How should you ship and store it?
从市场需求、库存现金、仓库和一件代发中选一个问题。Start with market demand, inventory cash, warehouses, or dropshipping.
我的库存概况My inventory overview项目、库存、现金和下一步Projects, stock, cash, and next action
全部物流工具All fulfillment tools库存、市场、社媒、路线Inventory, markets, social, and routes
库存投入决策Inventory investment
库存与现金计划Inventory and cash plan
区域信号Regional signals
兴趣、覆盖与供货地Interest, coverage, and supply
兴趣Interest
配送覆盖Offer coverage
一件代发Dropship route
正在读取地图...Loading map...
趋势观察Trend observation
社媒市场雷达Social market radar
履约路线Fulfillment routes
一件代发可能性Dropshipping possibilities
库存动作Inventory actions
| 产品Product | 需求Demand | 报价Offers | 库存覆盖Stock coverage | 陈旧数据Stale | 风险Risk | 动作Action |
|---|
选品决策流程PRODUCT DECISION WORKFLOW
选一个值得继续研究的商品Choose one product worth investigating
从“决策”开始。只有候选商品需要进一步调查时，才打开评论、利润或供应商证据。Start at Decide. Open reviews, economics, or supplier evidence only when a candidate needs more investigation.
等待加载Ready to load
数据覆盖与证据限制Data coverage & evidence limits需要时查看采集范围和真实性规则Open when you need collection scope or evidence rules
证据规则：Evidence rule:
排名不是销量，历史评论不能证明当前卖家表现，供应商目录条目也不是已核验工厂。样品、文件和现场证据仍需人工核验。
Rank is not sales, historical reviews do not prove current-seller performance, and directory records are not verified factories. Samples, documents, and facility evidence still require human verification.
正在读取评论覆盖…Loading review coverage…
第一步STEP ONE
今天先研究哪些商品？Which products should I investigate today?
每个候选只给一个建议动作。先处理最上面的项目。Each candidate gets one recommended action. Start at the top.
筛选候选商品Filter candidates
正在建立决策队列…Building the decision queue…
亚马逊公开商业证据OBSERVED AMAZON COMMERCIAL EVIDENCE
当前热卖商品与排名变化Current bestsellers and rank movement
畅销榜、24 小时飙升榜、新品榜、心愿榜与礼赠榜分开保存；“全榜重叠”仅按市场与 ASIN 连接，不合并名次。飙升榜位置不是销量，只保存亚马逊公开显示的 24 小时销售排名增幅和前次排名。其他榜单名次也不冒充销量。Best Sellers, Movers & Shakers, New Releases, Most Wished For, and Most Gifted remain separate; “All lists & overlap” joins only by marketplace and ASIN and does not merge ranks. A Movers & Shakers list position is not sales; Kouqiuqiu stores only Amazon’s visibly rendered 24-hour sales-rank gain and previous rank when available. No list rank is labeled as sales.
真实商品语料，不伪造搜索量OBSERVED PRODUCT CORPUS — NO FAKE SEARCH VOLUME
关键词与反查 ASINKeyword mining and reverse ASIN
关键词来自当前亚马逊热卖商品标题。需求指数基于排名覆盖和变化；竞争指数基于商品密度与评论门槛。它们不是搜索量、广告竞价或转化率。Keywords come from current Amazon bestseller titles. Demand uses rank coverage and movement; competition uses product density and the observed review barrier. These are not search volume, ad bids, or conversion rates.
亚马逊原生搜索语言AMAZON-NATIVE SEARCH LANGUAGE
实时自动补全关键词扩展Live autocomplete keyword expansion
按种子词和 A–Z 前缀采集亚马逊公开自动补全。排名表示当时的建议位置；覆盖表示该词出现于多少个查询变体。两者都不是搜索量、销量、点击份额或转化率。Collects Amazon's public autocomplete for the seed and bounded A–Z expansions. Rank is the observed suggestion position; coverage is the number of query variants that returned the term. Neither is search volume, sales, click share, or conversion.
选择单个市场、输入种子词，然后采集亚马逊词。Choose one marketplace, enter a seed, then discover Amazon terms.
持续采集CONTINUOUS AUTOCOMPLETE OBSERVATION
已保存的亚马逊种子词跟踪Saved Amazon autocomplete trackers
系统按计划重复采集公开自动补全并保留位置变化。自动种子词来自已观察的畅销榜、新品榜、心愿榜与礼赠榜标题，不代表搜索量。The scheduler repeats public autocomplete collection and retains position history. Automatic seeds come from observed Best Sellers, New Releases, Most Wished For, and Most Gifted titles and are not search-volume claims.
尚未保存自动补全跟踪。No scheduled autocomplete trackers yet.
公开搜索结果实测OBSERVED PUBLIC SEARCH PLACEMENT
真正的反查 ASIN：关键词排名观察Reverse ASIN: observed keyword placement
仅记录某个时间点亚马逊公开搜索第一页是否出现该 ASIN，以及自然/广告位置。未出现不等于完全没有排名；系统不会把它伪装成搜索量、流量份额、转化率或销量。This records only whether the ASIN appears on page one of Amazon's public search results at a timestamp, with organic versus sponsored position. Not found does not prove no ranking, and this is never labeled as search volume, traffic share, conversion, or sales.
输入已观察的 ASIN 开始实测。Enter an observed ASIN to run placement checks.
自动复查SCHEDULED PUBLIC CHECKS
已保存的关键词排名跟踪Saved keyword placement trackers
尚未保存跟踪目标。No scheduled placement trackers yet.
市场 × 品类基准MARKET × CATEGORY BENCHMARKS
细分市场机会与进入门槛Niche opportunity and entry barriers
品类基准分别保留畅销榜、新品榜、心愿榜和礼赠榜，并计算同一 ASIN 的跨榜重叠；该重叠不是订单量或销量。热卖排名历史、主币种价格区间、评论集中度、卖家身份覆盖和本站首次观察时间仍单独展示。首次观察不是上架日期；季节性只有在跨度至少 90 天且观察日足够时才会解锁。Category benchmarks keep Best Sellers, New Releases, Most Wished For, and Most Gifted separate and calculate same-ASIN overlap; overlap is not order count or sales. Bestseller-rank history, dominant-currency price bands, review concentration, seller-identity coverage, and first-observed timestamps remain separate. First observed is not a listing date, and seasonality unlocks only after at least a 90-day span with sufficient observed days.
跨市场相似商品CROSS-MARKET PRODUCT SIMILARITY
ASIN 与竞品基准ASIN and competitor benchmarks
相似度来自标题词、品类和市场重合，是待核实的竞品关系；不会伪装成亚马逊官方“相关商品”。Similarity comes from title terms, category, and market overlap. It is an inferred competitor relationship, not an Amazon-official related-product claim.
输入 ASIN 或商品词开始对比。Enter an ASIN or product phrase to build a competitor set.
排名、竞争、供应商与单位经济RANK, COMPETITION, SUPPLIERS & UNIT ECONOMICS
选品对比与利润压力测试Product comparison and profit stress test
供应商匹配来自商品词和公开目录能力，只是待核实线索。目录价格不当作正式报价；产品成本必须由你输入或通过询价确认。Supplier matches use shared product terms and public directory capabilities; they are leads to verify. Directory prices are not treated as quotes, and product cost must be entered or confirmed through an RFQ.
在亚马逊选品列表点击“对比”添加商品。Use Compare in Amazon research to add products.
买家原话、故障主题和实拍图BUYER WORDS, FAILURE THEMES & CUSTOMER PHOTOS
ASIN 评论与仿冒线索ASIN review and counterfeit clues
分析公开页面片段和公开研究语料中的历史商品评论。历史抽样偏向问题与实拍，不代表完整或有代表性的评论库；仿冒关键词是调查线索，不是判定，也不能归因于当前 Featured Offer 卖家。This analyzes visible public-page snippets and historical product reviews from public research corpora. The historical sample favors complaints and photos, so it is neither complete nor representative. Authenticity phrases are leads, not verdicts, and cannot be attributed to the current featured-offer seller.
质量历史与下一步Quality history and next step
购买前后的真实问题REAL QUESTIONS AROUND PURCHASE
买家担心什么？What did buyers worry about?
来自亚马逊 2019–2020 年 Apache-2.0 公开研究语料。问题与 ASIN 直接关联，但不是评论、当前需求或当前卖家证据。From Amazon's Apache-2.0 2019–2020 research corpus. Questions are linked directly to ASINs, but they are not reviews, current demand, or current-seller evidence.
重复问题主题Recurring problem themes
评论证据Review evidence
买家实拍图Customer photos
中国与零售商公开买家证据CHINA & RETAILER BUYER EVIDENCE
已采集评论与买家实拍库Collected reviews and customer-photo library
每条评论和图片都保留来源。平台“已购买”标签仍只是平台声明；只有明确绑定卖家商品的证据才能用于卖家风险判断。Every review and image retains its source. A platform purchase badge remains a platform claim, and only evidence explicitly tied to a seller listing can contribute to seller risk.
中国卖家级评论与实拍证据CHINA SELLER-SPECIFIC REVIEW & PHOTO EVIDENCE
卖家商品的真伪调查线索Authenticity investigation by seller listing
这里只有明确绑定卖家商品的评论才进入卖家档案。系统把卖家实拍与同款参考实拍比较；图片差异、低价或单条投诉都不能单独证明仿冒。话题页和品类页只扩充参考库，不能用于指控卖家。Only reviews explicitly tied to a seller listing enter these profiles. Seller-linked photos are compared with exact-product reference photos; photo differences, low price, or a single complaint cannot independently prove counterfeit. Topic and category pages expand the reference corpus but cannot be used to accuse a seller.
正在读取卖家证据…Loading seller evidence…
持久观察列表PERSISTENT OBSERVATION WATCHLIST
ASIN 排名与评论跟踪ASIN rank and review trackers
跟踪器保留 ASIN，即使它暂时离开热卖榜。缺失表示当前公开榜单没有观察到，不等于停产或下架。Trackers retain an ASIN even if it falls out of the observed bestseller pages. Missing means not currently observed; it does not prove discontinuation or delisting.
在亚马逊选品列表点击“跟踪”来添加商品。Use Track from the Amazon product list to add an ASIN.
公开观察证据，不是虚构销量OBSERVED PUBLIC EVIDENCE, NOT INVENTED SALES
亚马逊趋势、供应商与防伪雷达Amazon trend, sourcing, and authenticity radar
趋势分只组合当前亚马逊热卖排名、公开自动补全可见性、最多 30 次真实采集历史、跨市场覆盖与观察到的排名/评论变化。“首次观察”只表示商品第一次进入本站采集到的热卖榜语料，不是亚马逊上架日期。供应商来自 1688、义乌购和广交会公开目录的词组匹配，尚不等于真实工厂关系。评论中的假货措辞和买家图片仅为调查线索，不是仿冒定论。系统没有把这些数据冒充搜索量或销量。The trend score combines only current Amazon bestseller placement, public autocomplete visibility, up to 30 real collection runs, cross-market coverage, and observed rank/review movement. “First observed” means first seen in this collected bestseller corpus—not an Amazon listing-creation date. Supplier matches are token-matched public leads from 1688, YiwuGo, and Canton Fair—not proof of an exact factory relationship. Authenticity phrases and customer photos are investigation leads, not counterfeit findings. None of this is labeled as search or sales volume.
中国公开目录PUBLIC CHINA DIRECTORIES
供应商发现与核验缺口Supplier discovery and verification gaps
无自动发送NO AUTOMATIC SENDING
询价审批与回复证据RFQ approvals and response evidence
采集透明度ACQUISITION TRANSPARENCY
供应商数据源状态Supplier source health
被反爬挑战阻止的来源会明确标记为 blocked；系统不会绕过挑战、伪造结果或用无关聚合站填数。Sources stopped by an access challenge are explicitly marked blocked. The system does not bypass challenges, fabricate records, or pad results with unrelated aggregators.