· TagPeek
· TagPeek
The 635px first-photo rule
635px 首图规则
- Treat the first photo as the catalog cover — aim ~635px on both sides. 把首图当目录封面——宽高都尽量够约 635px。
- 635px and 2000px image-zoom are separate checks; pass both if you need zoom. 635px 与 2000px 图片放大是两道关；需要放大就两道都过。
- After swapping the hero, re-audit the live page once cache catches up. 换首图后等缓存跟上，再审计一次线上页。
I treat my first listing photo as the catalog cover. Etsy repeats it in search results, favorites grids, and many off-Etsy previews. It is not always photo slot one in my creative story, but it is the hero the platform publishes size floors for.
我把第一张商品页照片当目录封面。Etsy 会在搜索结果、收藏网格以及许多站外预览里反复用它。它不一定是我创意叙事里的第一槽，但平台为它单独公布了尺寸底线。
TagPeek cites a 635 pixel minimum width and height for the first photo in Rule 02, matching Etsy Help language about search visibility. Falling short does not always remove the listing, but next to sharper competitors you can look soft and lose clicks.
TagPeek 在规则 02 中引用首图宽高各至少约 635 像素，并与 Etsy Help 关于搜索可见性的表述一致。不达标不一定下架商品页，但在已达标的竞品旁边，你可能显得更糊、更吃亏。
Why 635px is not the same as zoom
为什么 635px 不等于放大
The 635px rule is separate from hover-zoom’s 2000px shortest-side requirement. An 800px square might pass the first-photo check while keeping zoom disabled. A 2500px wide panorama might enable zoom but crop awkwardly in 1:1 search tiles if the subject sits outside the safe zone.
635px 规则与图片放大的 2000px 最短边要求是分开的。800px 正方形可能通过首图检查，却仍让放大关闭。2500px 宽的全景图可能开启放大，但若主体落在安全区外，1:1 搜索瓷砖里裁切会很尴尬。
I use the detector’s crop previews to see how Etsy trims the same file across 1:1, 4:5, 4:3, and 5:4 frames. Orange safe zones mark areas likely to survive grid crops. Subject placement matters as much as raw resolution.
我用检测器的裁切预览，看同一文件在 1:1、4:5、4:3、5:4 框里如何被裁。橙色安全区标出更可能在网格裁切中幸存的区域。主体摆位与原始分辨率同等重要。
How I keep heroes search-ready
我怎么让首图在搜索里站得住
Mobile-first photography can accidentally shrink the short side. If you export vertical stills to a square canvas with wide margins, the product pixels might not fill enough of the frame to look sharp when shrunk.
以手机为主的拍摄，可能无意中压短短边。若把竖版静帧导出到带大留白的正方形画布，产品像素填不满画面，缩小后就会显得不锐。
Consistency across listings helps buyers recognize my shop. I keep hero exports on a shared preset: sRGB, enough short side for zoom if detail matters, and 635px or larger on both dimensions for slot one.
跨商品页保持一致，有助于买家认出我的店。首图导出共用一套预设：sRGB；若细节重要则短边足以开放大；第一槽位宽高都不低于 635px。
TagPeek’s listing health score bundles first-photo signals with tag and title checks. None of those numbers predict sales. They summarize how closely your live listing matches published requirements you can verify yourself.
TagPeek 的商品页健康分会把首图信号与标签、标题检查打包。这些数字都不预测销量。它们只概括线上商品页与可独立核实的公开要求有多接近。
When I swap the first photo after launch, I re-run the extension audit. Etsy reprocesses uploads; page caches can take time to catch up. A passing local file plus a failing live read usually means I uploaded the wrong file, or old cache is still showing.
上线后若更换首图，我重跑扩展审计。Etsy 会重新处理上传；页面缓存可能需要时间。本地文件通过而线上读取失败，通常意味着传错了文件，或旧缓存仍在显示。
I pair this rule with transparency and watermark self-checks in the detector. Some programs exclude watermarked heroes even when pixels are ample.
我把本规则与检测器里的透明通道、水印自检一起看。即便像素充足，带水印的首图也可能让某些计划失去资格。
If I keep separate photos for Pinterest or Instagram, I still validate the Etsy hero on its own. Cross-posting compressed social exports is a common way shops dip under 635px without noticing.
若我为 Pinterest 或 Instagram 另备照片，仍要单独校验 Etsy 首图。把压缩过的社交导出直接跨发，是店铺不知不觉掉到 635px 以下的常见原因。
I document the export recipe the detector generates in our shop notes. New helpers can paste the same dimensions and quality settings instead of guessing.
我把检测器生成的导出配方写进店铺笔记。新人可以直接粘贴同样的尺寸与质量设置，而不是靠猜。
Square mockups with too much padding shrink the product share of the frame. The 635px rule measures the file, but human shoppers still punish tiny subjects surrounded by empty margin. Crop tighter before chasing more pixels.
留白过多的正方形样机，会缩小产品在画面中的占比。635px 规则量的是文件，但真人买家仍会嫌弃被空边包围的小主体。先裁紧，再追像素。
If I A/B test heroes, I change one variable at a time. Swapping photos and tags simultaneously makes it impossible to know which edit moved my TagPeek score.
若做首图 A/B 测试，我一次只改一个变量。同时换照片与标签，就无法知道是哪次编辑推动了 TagPeek 分数。
Handmade variance means two finished mugs may photograph differently. I standardize lighting and distance so the hero represents the typical item, then verify pixels on that representative shot.
手作存在差异，两只成品马克杯拍出来可能不同。我标准化灯光与拍摄距离，让首图代表典型货品，再在那张代表照上核对像素。
Bundles with multiple items need heroes that read at thumbnail scale. A collage of four tiny products may technically pass 635px while failing the squint test in search.
多件捆绑商品的首图，需要在缩略图尺寸下仍可读。四件产品拼成的拼贴可能技术上过 635px，却过不了搜索里的「眯眼测试」。
If I migrate from another platform, I re-export heroes instead of reusing compressed marketplace thumbnails. Those imports often arrive already under the first-photo floor.
若从其他平台迁店，我重新导出首图，不要复用已压缩的市场缩略图。这类导入往往一到货就已低于首图底线。
Listing videos do not replace the still hero for grid crops. I treat video as supplemental while the first photo carries search square duty.
商品页视频不能替代静帧首图承担网格裁切。视频当作补充；搜索方块仍由首图负责。
Seasonal overlays on heroes — holiday badges, sale ribbons — should stay inside the safe zone in crop previews. A badge clipped in search can look unprofessional even when raw pixels pass.
首图上的季节性贴片——节日徽章、促销丝带——应落在裁切预览的安全区内。搜索里被裁掉的徽章，即便原始像素过关，也显得不专业。
If you use AI upscalers, inspect edges for halos before upload. Halos survive 635px checks while failing buyer trust.
若使用 AI 放大，上传前检查边缘是否有光晕。光晕能过 635px 检查，却过不了买家信任。
I compare my first photo to competitors in the same search query on desktop. I am not copying style; I am calibrating whether my thumbnail competes at card size.
在桌面端同一搜索词下，我把首图与竞品并排看。不是抄风格，而是校准缩略图在卡片尺寸下是否站得住。