使用集合让一切井井有条
根据您的偏好保存内容并对其进行分类。
实施灵活抽样时需遵循的常规指南
为了更好地了解抽样方面的更改可能会对 Google 用户和发布商的订阅模式产生怎样的影响,我们和发布合作伙伴共同开发了一系列实验。从这些实验中,我们了解到:即便是对当前抽样等级的细微更改,也可能会有损用户体验,并可能会无意中影响报道在 Google 搜索中的排名,因为用户访问权限受到了限制。
我们建议采用的抽样类型有两种:第一种是计量供给 - 先为用户提供一定量的免费文章供浏览,然后要求用户订阅或登录,之后就会开始显示付费墙;第二种是导入式抽样 - 仅提供一篇文章的部分内容,而不显示完整文章。
我们建议发布商使用不同数量的抽样内容谨慎地实验。
以下是实施灵活抽样时需遵循的一些常规指南。
计量供给
一般来说,对于计量供给,我们认为按月计量(而非按日计量)能提供更高的灵活性和更安全的测试环境。样本数量从一个整数值变成另一个整数值(例如,每月 10 个样本而不是每日 3 个样本),给用户所带来的影响不会那么显著。按月计量还具有以下优势:可将付费墙主要显示给互动程度最高的用户(即最有可能订阅的用户),同时也可让新用户和互动程度较低的用户在遇到付费墙之前有机会了解一下内容的价值。(在这种情况下,“付费墙”既可指必须订阅才能访问内容的障碍设置,也可指只需注册即可访问内容的障碍设置。)
多少内容?
对于各个行业,没有一个通用的最佳抽样值。不过,对于大多数每日新闻发布商,我们希望该值的范围介于每月每位用户 6 到 10 篇报道之间。我们认为,大多数发布商都能在该范围内找到一个可实现下述目标的值:既能为新的潜在订阅者提供良好的用户体验,又能提高互动程度最高的用户的转化率。
在探索新的方法时,建议您一开始每月向 Google 搜索用户提供 10 篇免费报道,然后以此值为基础不断调整。我们将具体数值的决定权交给各个发布商,因为他们最了解自己业务的独特需求。我们建议发布商先分析已遇到付费墙的搜索用户所占的百分比,然后再选择一个能达到近似结果的每月数值。确信自己已打好稳固的基础之后,您可以随时降低该值。
导入式抽样
除了计量供给之外,某些发布商会在计量供给用尽后,在付费墙的“首屏”中展示一篇报道的前几句内容。我们认为这是一种很棒的做法。与完全不让用户看到任何内容的网页相比,通过展示报道的开头部分,发布商可让用户了解该内容的价值,因此对用户更有意义。导入式抽样还会激发用户的好奇心,使他们非常想知道报道的下文,这可能会有助于促成转化。
如何更改
发布商需要尝试不同的抽样值,确定这些值对引荐流量和转化的影响。
请记住,我们的用户研究已表明,如果用户在只体验了少量内容时就被要求进行订阅,他们对该产品的兴趣就会大幅降低。我们的分析结果表明,如果付费墙在超过 10% 的时间里都是处于显示状态(这通常意味着约有 3% 的受众已遇到付费墙),整体的用户满意度就会开始明显下降。我们建议您切勿轻易接近这个限值,否则用户尚未对您的内容价值产生信心,就开始离开了。
如果发布商拥有更先进的技术资源,建议将精力专注于互动程度较高的细分受众群内的具体用户。通过确定总会用尽每月配额的用户,发布商便可有针对性地降低面向这些用户提供的样本限额,将他们视为目标客户;同时,通过向其他用户分配更多可供自由使用的限额,发布商还可避免令整体的用户行为和满意度受到不利影响的风险。
如何标示付费内容
请使用结构化数据标记付费内容,帮助 Google 区分付费内容与伪装真实内容的做法;伪装真实内容是指向 Googlebot 提供的内容不同于向用户提供的内容。
如果您不希望在提供内容时浏览器能够访问相应内容,请选择一种不会向浏览器提供付费内容的付费墙实现方式。
详细了解如何使用结构化数据标示付费内容,并参阅关于使用 JavaScript 实现付费内容的指南。
如未另行说明,那么本页面中的内容已根据知识共享署名 4.0 许可获得了许可,并且代码示例已根据 Apache 2.0 许可获得了许可。有关详情,请参阅 Google 开发者网站政策。Java 是 Oracle 和/或其关联公司的注册商标。
最后更新时间 (UTC):2025-09-01。
[null,null,["最后更新时间 (UTC):2025-09-01。"],[[["\u003cp\u003eFlexible sampling, including metering and lead-in, allows publishers to test the impact of paywalls on user experience and subscriptions.\u003c/p\u003e\n"],["\u003cp\u003eMetering, offering a monthly quota of free articles, is recommended over daily metering for greater flexibility and user engagement.\u003c/p\u003e\n"],["\u003cp\u003ePublishers should experiment with sampling, starting with around 6-10 free articles monthly and adjusting based on user behavior and conversion rates.\u003c/p\u003e\n"],["\u003cp\u003eShowing a lead-in, or a portion of the article before the paywall, can enhance user experience and encourage subscriptions.\u003c/p\u003e\n"],["\u003cp\u003eExcessive paywall encounters can negatively impact user satisfaction, so publishers should monitor and adjust sampling strategies accordingly.\u003c/p\u003e\n"]]],["Publishers should experiment with metering and lead-in sampling for paywalled content. Metering, preferably monthly, grants a set number of articles before a paywall appears; 6-10 monthly articles is recommended, starting at 10. Lead-in shows a portion of the article above the paywall. Cautious experimentation is crucial as excessive paywalls (over 10% of user interactions) reduce user satisfaction. Publishers can target engaged users with stricter metering and should indicate paywalled content using structured data.\n"],null,["Flexible Sampling general guidance\n\nIn order to better understand the potential impact of sampling changes on Google users\nand publishers' subscription models, we developed a series of experiments in cooperation with\nour publishing partners. From these experiments we learned that even minor changes to the\ncurrent sampling levels could degrade user experience and, as user access is restricted,\nunintentionally impact article ranking in Google Search.\n\nThere are two types of sampling we advise: **metering** , which provides users\nwith a quota of articles to consume before requiring users to subscribe or log in, after\nwhich paywalls will start appearing; and **lead-in**, which offers a portion of\nan article's content without it being shown in full.\n\nWe encourage publishers to experiment cautiously with different amounts of sampling.\nHere is some general guidance for implementing flexible sampling.\n\nMetering\n\nIn general, we think that monthly, rather than daily metering provides more flexibility and\na safer environment for testing. The user impact of changing from one integer value to the\nnext is less significant at, say, 10 monthly samples than at 3 daily samples. Monthly\nmetering also has the advantage of focusing paywall views on your most engaged users, who are\nthose most likely to subscribe, while allowing your newer and less engaged users to become\nacquainted with the value of your content before experiencing a paywall. (\"Paywall,\" in this\ncontext, applies equally to barriers that require either subscription or merely registration\nfor content access.)\n\nHow much content?\n\nThere is no single value for optimal sampling across different businesses. However,\nfor most daily news publishers, we expect the value to fall between 6 and 10 articles per\nuser per month. We think most publishers will find a number in that range that preserves a\ngood user experience for new potential subscribers while driving conversion opportunities\namong the most engaged users.\n\nAs a starting point for your explorations, we encourage you to provide 10 articles per month\nto Google search users and iterate from there. We leave the exact number to the discretion of\nindividual publishers, who are best positioned to understand the particular demands of their\nbusinesses. We encourage publishers to analyze the current percentage of search users who land\non their paywalls, and select a monthly number that achieves a similar result. You can always\nlower the value later, after you have some confidence that you are on a stable footing.\n\nLead-in\n\nIn addition to metering, some publishers show the first few sentences of an article \"above\nthe fold\" of their paywall after the meter has run out. We think this is a good practice. By\nexposing the article lead, publishers can let users experience\nthe value of the content and so provide more value to the user than a page with completely\nblocked content. Lead-in also generates user curiosity about how article continues, which\nmay assist in conversion.\n\nMaking Changes\n\nPublishers will want to experiment with different sampling values to determine their\neffect on referral traffic and conversion.\n\nBear in mind that our user studies have shown that when users who have experienced only a\nsmall amount of content are required to subscribe, their interest in the product diminishes\ngreatly. Our analysis shows that general user satisfaction starts to degrade significantly\nwhen paywalls are shown more than 10% of the time (which generally means that about 3% of\nthe audience has been exposed to the paywall). We recommend caution in approaching that limit,\nbecause you may start to alienate users who have not yet become convinced of the value of your\ncontent.\n\nPublishers with more advanced technical resources may want to focus their efforts more\nnarrowly on those specific users in the engaged segment. By identifying users who\nconsistently use up the monthly allotment, publishers could then target them by reducing the\nsample allowance for that audience specifically, and, by allowing more liberal\nconsumption for other users, reduce the risk that overall user behavior and satisfaction is\ndegraded.\n\nHow to indicate paywalled content\n\nEnclose paywalled content with structured data in order to help Google\ndifferentiate paywalled content from the practice of [cloaking](/search/docs/advanced/guidelines/cloaking),\nwhere the content served to Googlebot is different from the content served to users.\nIf you don't want the content to be accessible to the browser at the time of serving, choose a\npaywall implementation that doesn't supply the paywalled content to the browser.\n\nLearn more about how to [indicate paywalled\ncontent with structured data](/search/docs/appearance/structured-data/paywalled-content) and refer to our [guidance on using JavaScript to implement paywalled content](/search/docs/crawling-indexing/javascript/fix-search-javascript#paywall)."]]