[{"data":1,"prerenderedAt":315},["ShallowReactive",2],{"blog-cms-migration-url-template":3,"blog-prev-cms-migration-url-template":102,"blog-next-cms-migration-url-template":208},{"id":4,"title":5,"author":6,"body":7,"canonical":86,"category":87,"cover":86,"date":88,"description":89,"draft":90,"extension":91,"keywords":86,"meta":92,"navigation":93,"noindex":90,"ogImage":86,"path":94,"seo":95,"seoDescription":86,"seoTitle":86,"stem":96,"tags":97,"updated":86,"__hash__":101},"blog\u002Fblog\u002Fcms-migration-url-template.md","换内容管理系统不掉排名，URL 和模板先对齐","元析智能",{"type":8,"value":9,"toc":75},"minimark",[10,14,19,31,35,38,41,45,48,51,54,57,60,63,66,69,72],[11,12,13],"p",{},"CMS 只是发布工具，掉不掉排名看迁移过程有没有保住三样东西：URL、模板结构、渲染方式。这三样任何一样变了，都要当成一次站改来对待，按迁移的规矩做，别当成\"换个后台\"。",[15,16,18],"h2",{"id":17},"一url-必须逐条映射","一、URL 必须逐条映射",[11,20,21,22,26,27,30],{},"旧 CMS 的固定链接结构是定死的，新 CMS 默认结构常常不同。别用自动规则猜，逐条映射一遍。我们见过最惨的一次，迁移脚本把文章 URL 从 ",[23,24,25],"code",{},"\u002Fblog\u002Fpost-name"," 改成了 ",[23,28,29],{},"\u002Fblog\u002F2026\u002Fpost-name","，2000 篇文章全部换地址，301 也没做，两个月自然流量掉了一半。",[15,32,34],{"id":33},"二模板结构要对齐","二、模板结构要对齐",[11,36,37],{},"H1、正文、内链、面包屑、结构化数据的位置要一致。换主题之后最常见的问题：正文被塞进多个嵌套容器，正文长度看起来没变，但首屏只剩标题和目录；H1 被做成图片；面包屑从模板里消失。",[11,39,40],{},"2026 年 3 月一个 2000 页的站从 WordPress 迁到自研 CMS，URL 没变，但新模板把正文放进折叠组件。30 天后收录掉了 40%，修复只用了两周。",[15,42,44],{"id":43},"三渲染方式要确认","三、渲染方式要确认",[11,46,47],{},"新 CMS 自带的主题可能用 JS 渲染列表页和文章页。旧的如果是静态生成的，新旧渲染方式不一致，收录表现会差一截。迁移前用 curl 各拉一遍，对比初始 HTML。",[15,49,50],{"id":50},"对照检查表",[11,52,53],{},"迁移前爬旧站，记录每个页面的 title、H1、正文长度、内链数、结构化数据类型、状态码。迁移后用同样的规则爬新站，逐项对比。我们的脚本会输出差异页列表，按流量排序，先修高流量的。",[15,55,56],{"id":56},"迁移的时间窗口",[11,58,59],{},"选在流量低谷，比如季度末的周末。迁移前两周冻结内容改版，前一周把所有 URL 清单、重定向规则、新站爬虫报告都准备好。上线后一周内每天看日志和索引报告，异常当天处理。",[11,61,62],{},"我们给客户排的时间表：第 1 周准备，第 2 周上线，第 3 到 6 周观察，第 7 周复盘。整个过程 SEO 要参与，不能只在最后一天收到一个链接。",[15,64,65],{"id":65},"四个常见坑",[11,67,68],{},"旧站的重定向插件没导；标签和分类页的 URL 全变了；图片路径变了导致 404；RSS 和 sitemap 地址变了，订阅用户和蜘蛛都断了线。每一条都能在迁移前检查出来，只是没人愿意花那半天。",[15,70,71],{"id":71},"立场",[11,73,74],{},"换 CMS 之前先问一句：新站谁负责 SEO 验收。没有验收环节的迁移就是赌博，赌注是你积累了几年的收录和排名。",{"title":76,"searchDepth":77,"depth":77,"links":78},"",2,[79,80,81,82,83,84,85],{"id":17,"depth":77,"text":18},{"id":33,"depth":77,"text":34},{"id":43,"depth":77,"text":44},{"id":50,"depth":77,"text":50},{"id":56,"depth":77,"text":56},{"id":65,"depth":77,"text":65},{"id":71,"depth":77,"text":71},null,"技术SEO","2026-06-21","从 WordPress 换到自研 CMS、从自研换到 Headless，我们做过 4 次。掉排名基本不是 CMS 本身的问题，而是迁移时没对齐 URL、模板和渲染。附一份迁移前后对照检查表。",false,"md",{},true,"\u002Fblog\u002Fcms-migration-url-template",{"title":5,"description":89},"blog\u002Fcms-migration-url-template",[98,99,100],"网站迁移","CMS","URL设计","zjkfN2lakH5QDTYAm6Lek9VCameqWBZoIj09nscSziM",{"id":103,"title":104,"author":6,"body":105,"canonical":86,"category":196,"cover":86,"date":197,"description":198,"draft":90,"extension":91,"keywords":86,"meta":199,"navigation":93,"noindex":90,"ogImage":86,"path":200,"seo":201,"seoDescription":86,"seoTitle":86,"stem":202,"tags":203,"updated":86,"__hash__":207},"blog\u002Fblog\u002Ftopic-selection-meeting.md","选题会开成聊天会，问题出在没带数据",{"type":8,"value":106,"toc":188},[107,110,113,116,119,122,125,128,131,134,137,140,143,146,149,152,155,158,161,164,167,170,173,176,179,182,185],[11,108,109],{},"2026 年 6 月我们复盘了上半年的选题会，效率最高的一次三十分钟出了八个直接能写的选题，最差的一次开了一个半小时，最后只定了一个。差别不在人，在会前有没有带数据。",[15,111,112],{"id":112},"会前准备三份数据",[11,114,115],{},"第一份，GSC 里已经有曝光但我们没有专门页面的词。这类词是现成的需求，写一篇文章就能接住。",[11,117,118],{},"第二份，竞品最近三个月新发的内容清单，看他们写了什么、哪些有互动。不是照抄，是看有没有我们漏掉的角度。",[11,120,121],{},"第三份，销售和客服最近一个月的真实问题。客户问的问题，就是最好的选题来源。",[11,123,124],{},"这三份数据准备起来两个小时，但能让会议从“我觉得”变成“数据显示”。",[15,126,127],{"id":127},"会上只回答三个问题",[11,129,130],{},"这篇写给谁。具体到角色，比如“正在选型的后端负责人”，不是“开发者”。",[11,132,133],{},"解决什么问题。一句话能说清，说不清就是选题太大。",[11,135,136],{},"凭什么我们能写。我们有实测数据、有客户案例、有踩坑经验，还是只能查资料拼一篇。第三种就别写。",[11,138,139],{},"三个问题答完，选题要么通过，要么砍掉，不讨论第三遍。",[15,141,142],{"id":142},"浪费时间的四种讨论",[11,144,145],{},"讨论标题。标题是写完之后的事，会上定标题只会陷入审美争论。",[11,147,148],{},"争论搜索量。工具给的数字前面已经看过，会上再争没有意义，而且软件的词量普遍偏小，争也争不出结果。",[11,150,151],{},"没有决策人。选题会必须有最终拍板的人，否则每个选题都要“再想想”。",[11,153,154],{},"一次定二十个选题。定得越多，写得越少。我们现在一次最多定八个，对应两周的产量。",[15,156,157],{"id":157},"会后怎么落地",[11,159,160],{},"每个选题填一张卡：标题方向、目标词、写给谁、核心观点、需要的素材、负责人、截止日期。",[11,162,163],{},"卡片填不出来的选题，说明还没想清楚，退回重想。",[11,165,166],{},"两周后复盘一次：写出来的效果怎么样，没写出来的卡在哪。这个动作比会上的讨论有用。",[15,168,169],{"id":169},"一个可量化的小指标",[11,171,172],{},"我们看选题质量不看阅读量，看这篇文章上线四周后有没有拿到目标词的曝光。有曝光说明选题命中了需求，没有就是自嗨。",[11,174,175],{},"这个指标比阅读量诚实，因为阅读量可以靠渠道推。",[15,177,178],{"id":178},"我的看法",[11,180,181],{},"选题会不是头脑风暴。头脑风暴适合找方向，选题会适合做减法。",[11,183,184],{},"我们现在的会固定三十分钟，前二十分钟过卡片，后十分钟排优先级。超时就停，剩下的下次再说。",[11,186,187],{},"写得出来比想得漂亮重要。一个能落地的普通选题，胜过十个停在会议纪要里的好想法。",{"title":76,"searchDepth":77,"depth":77,"links":189},[190,191,192,193,194,195],{"id":112,"depth":77,"text":112},{"id":127,"depth":77,"text":127},{"id":142,"depth":77,"text":142},{"id":157,"depth":77,"text":157},{"id":169,"depth":77,"text":169},{"id":178,"depth":77,"text":178},"内容运营","2026-06-20","三十分钟出八个能落地的选题，前提是会前准备三份数据。这篇写选题会只回答哪三个问题、哪些讨论纯属浪费、会后选题卡怎么填，开完就能用。",{},"\u002Fblog\u002Ftopic-selection-meeting",{"title":104,"description":198},"blog\u002Ftopic-selection-meeting",[204,205,206],"选题","内容流程","团队协作","7NWtozYEVdlC3vvhVqM4ucRZnfWwsQk9JyE1-fVvbL4",{"id":209,"title":210,"author":6,"body":211,"canonical":86,"category":87,"cover":86,"date":304,"description":305,"draft":90,"extension":91,"keywords":86,"meta":306,"navigation":93,"noindex":90,"ogImage":86,"path":307,"seo":308,"seoDescription":86,"seoTitle":86,"stem":309,"tags":310,"updated":86,"__hash__":314},"blog\u002Fblog\u002Ffaceted-navigation-crawl.md","筛选参数把抓取预算吃光了：电商站的 5 个处理规则",{"type":8,"value":212,"toc":294},[213,216,219,222,225,229,232,235,239,249,252,256,259,262,266,269,272,276,279,282,285,288,291],[11,214,215],{},"一个 B2B 电商站的筛选组合生成了 6 万多个 URL，抓取预算的 62% 花在参数页上，核心产品页一周才被抓一次。我们用 5 条规则把参数页从 6 万压到 800。",[15,217,218],{"id":218},"先算清楚参数页有多少",[11,220,221],{},"三个筛选维度各 10 个值，组合起来理论上是 1000 种；如果再叠加排序、页码、视图模式，实际 URL 能到几万个。",[11,223,224],{},"查法：从日志里把带问号的 URL 全提出来，按参数名分组，数每个参数组合的抓取次数和收录数。我们那个客户，日志里 62% 的抓取请求落在参数页，而这些页面的收录率只有 8%。",[15,226,228],{"id":227},"规则一只保留有搜索需求的组合","规则一：只保留有搜索需求的组合",[11,230,231],{},"不是所有筛选组合都值得收录。用关键词工具查\"产品词 + 筛选值\"有没有搜索量，有量的保留，没量的不收录。",[11,233,234],{},"那个客户 6 万多个组合里，真正有搜索需求的只有 800 个左右。",[15,236,238],{"id":237},"规则二参数顺序统一去掉无意义参数","规则二：参数顺序统一，去掉无意义参数",[11,240,241,244,245,248],{},[23,242,243],{},"?color=red&size=1"," 和 ",[23,246,247],{},"?size=1&color=red"," 是两个 URL，内容一样。排序参数、追踪参数、会话参数，全部规范化或者屏蔽。",[11,250,251],{},"做法：服务端对参数排序，追踪参数用 canonical 合并，视图模式参数直接不走 URL。",[15,253,255],{"id":254},"规则三没有搜索量的组合用-robots-屏蔽","规则三：没有搜索量的组合用 robots 屏蔽",[11,257,258],{},"屏蔽的是\"组合页\"，不是\"单个筛选值页\"。比如颜色筛选可以收录，颜色 + 尺码 + 价格区间的三重叠组合就屏蔽。",[11,260,261],{},"robots 屏蔽的页面不会被抓取，也就不会浪费预算。代价是这些页面也拿不到排名，但它们本来也没有排名。",[15,263,265],{"id":264},"规则四筛选页要有独立价值","规则四：筛选页要有独立价值",[11,267,268],{},"保留下来的 800 个筛选页，不能只是列表换个顺序。我们给每个页面加了筛选条件说明、适用场景、相关产品推荐，让它有独立的内容价值。",[11,270,271],{},"这一步决定了收录之后能不能排上去。只靠参数组合的页面，收录了也是零展现。",[15,273,275],{"id":274},"规则五内链只指向有价值的筛选页","规则五：内链只指向有价值的筛选页",[11,277,278],{},"导航和列表页里的筛选链接，只链到保留下来的 800 个页面。无价值的组合不生成可点击链接，减少蜘蛛的发现路径。",[15,280,281],{"id":281},"结果",[11,283,284],{},"调整从 2026 年 1 月开始，做了 6 周。第 8 周看数据：抓取请求里参数页占比从 62% 降到 9%，核心产品页的平均抓取间隔从 7 天缩到 1.5 天，有排名的关键词从 1200 个涨到 3100 个。",[11,286,287],{},"询盘量第 10 周开始涨，两个月后比调整前高 40%。",[15,289,290],{"id":290},"一句判断",[11,292,293],{},"筛选页不是\"要不要收录\"的问题，是\"哪些组合值得收录\"的问题。先算清楚数量和需求，再决定屏蔽还是保留，别一刀切全屏蔽——那会丢掉真正有搜索量的长尾。",{"title":76,"searchDepth":77,"depth":77,"links":295},[296,297,298,299,300,301,302,303],{"id":218,"depth":77,"text":218},{"id":227,"depth":77,"text":228},{"id":237,"depth":77,"text":238},{"id":254,"depth":77,"text":255},{"id":264,"depth":77,"text":265},{"id":274,"depth":77,"text":275},{"id":281,"depth":77,"text":281},{"id":290,"depth":77,"text":290},"2026-06-22","一个 B2B 电商站的筛选组合生成了 6 万多个 URL，抓取预算的 62% 花在参数页上，核心产品页一周才被抓一次。我们用 5 条规则把参数页从 6 万压到 800，收录率和询盘都回来了。照着做能把参数页数量压掉 90%，还能把收录率拉回来。",{},"\u002Fblog\u002Ffaceted-navigation-crawl",{"title":210,"description":305},"blog\u002Ffaceted-navigation-crawl",[311,312,313],"筛选参数","抓取预算","电商SEO","Rci4uxEa3PhEdxr9-z2RKNwbjuuZvV72GCr1U7bkajk",1794088104541]