本政策说明 GEO Tag 在提供网站访客分析服务过程中，收集哪些信息、如何处理、保留多久，以及你可以行使哪些权利。文中所写的每一项均与产品实际行为一致——没有写"可选功能"却未实现的情况。 This policy explains what information GEO Tag collects while providing website visitor analytics, how it is processed, how long it is kept, and what rights you can exercise. Every statement here matches what the product actually does — no capabilities are described that are not implemented.
生效日期：2026 年 7 月 31 日 · 最近更新：2026 年 7 月 31 日 Effective: 31 July 2026 · Last updated: 31 July 2026
GEO Tag 是一款面向企业的网站分析工具。使用本产品的网站经营者（我们的客户）把一段代码装在自己的网站上，我们代其采集与统计该网站的访客行为数据。 GEO Tag is a business analytics tool. The website operator (our customer) installs a snippet on their own site, and we collect and aggregate that site's visitor behaviour data on their behalf.
| 角色Party | 是谁Who | 法律地位Legal role |
|---|---|---|
| 网站经营者Website operator | 安装了本产品代码的网站方The site that installed our snippet | 个人信息处理者。决定为什么采集、采集什么，并须自行向其访客履行告知与取得同意的义务 Data controller. Decides why and what is collected, and is responsible for notifying its own visitors and obtaining consent |
| 我们Us | 上海超响应数字科技有限公司 | 受托处理者。仅按客户指示处理数据，不将其用于自身目的 Data processor. Processes data only on the customer's instructions, never for our own purposes |
这个划分很重要：如果你是某个网站的访客、想了解自己的数据为何被采集，应当先联系那个网站；我们无法代其回答采集目的，也不掌握你在该网站的账号身份。 This distinction matters: if you are a visitor to a site and want to know why your data is collected, contact that site first. We cannot answer for their purposes and we do not hold your account identity on their site.
以下是数据库中实际存储的全部字段，没有隐藏项： These are all the fields actually stored in the database — there are no hidden ones:
| 类别Category | 具体内容Contents |
|---|---|
| 标识Identifiers | 访客 ID、会话 ID、是否新访客。访客 ID 是随机生成的匿名字符串，存在访客自己浏览器里，不含任何真实身份信息，也不与任何账号绑定 Visitor ID, session ID, new-visitor flag. The visitor ID is a randomly generated anonymous string stored in the visitor's own browser; it contains no real identity and is not linked to any account |
| 行为Behaviour | 事件名称、发生时间、页面路径Event name, timestamp, page path |
| 来源归因Attribution | 来路引擎（如豆包 / DeepSeek / 通义千问）、来源标识、归因置信度；首次来源与本次会话来源分别记录。若网站经营者启用了可选的 Server-Timing 回传，还会记录一个到达方式枚举值（cross-site / same-site / same-origin / none，来自浏览器的 Sec-Fetch-Site）——它只表明"是否从别的站点跳来"，不包含来源网址、不含任何可识别个人的信息
Referring engine (e.g. Doubao / DeepSeek / Qwen), source marker, attribution confidence; first-touch and session source recorded separately. Where the site operator has enabled the optional Server-Timing echo, one arrival type enum is also recorded (cross-site / same-site / same-origin / none, from the browser's Sec-Fetch-Site) — it states only whether the visit came from another site, and carries no source URL and nothing personally identifying |
| 环境Environment | 设备类型（桌面/移动/平板三档）、国家、省、市，以及浏览器随每个请求自动发送的 User-Agent 字符串（含浏览器与操作系统版本，以及在 App 内打开时该 App 的标识） Device type (desktop / mobile / tablet), country, province, city, plus the User-Agent string the browser sends automatically with every request (browser and OS version, and the app identifier when the page is opened inside an app) |
| 网址参数URL parameters | 来路网址（referrer），以及落地网址上的查询串与锚点原文（各截断至 200 字符）。敏感参数的值在发送前即在访客浏览器内被替换为 [redacted]，包括令牌、密码、密钥、邮箱、手机号、会话与授权码等——参数名保留，参数值不出浏览器
The referrer, plus the raw query string and fragment of the landing URL (each capped at 200 characters). Values of sensitive parameters are replaced with [redacted] inside the visitor's browser before anything is sent — tokens, passwords, keys, emails, phone numbers, session and authorisation codes. Parameter names are kept; their values never leave the browser |
| 爬虫标记Bot flag | 是否为爬虫及其名称（爬虫流量单独统计，不计入访客） Whether the hit is a bot and its name (bot traffic is reported separately, not counted as visitors) |
| 自定义事件参数Custom event parameters | 由网站经营者自行决定传入，例如转化名称、金额。我们原样存储，不做解析 Decided entirely by the website operator, e.g. conversion name or amount. We store them verbatim without interpretation |
致网站经营者：自定义事件参数是你唯一可能把个人信息（姓名、手机号、订单号、邮箱等）传进来的通道。请不要这么做——那既超出本产品的设计用途，也会让你自己承担额外的合规责任。 To website operators: custom event parameters are the only channel through which personal information (names, phone numbers, order IDs, emails) could reach us. Please do not put such data there — it is outside this product's intended use and creates compliance obligations you would bear.
控制台账号仅存：邮箱、称呼、角色、状态、密码的 bcrypt 哈希值（不存明文密码）、创建时间、最近登录时间。管理员的写操作会留痕，其中包含操作者邮箱与操作时的 IP，用于安全审计。 Console accounts store only: email, display name, role, status, a bcrypt hash of the password (never the plaintext), creation time, and last login time. Administrator write actions are logged, including the operator's email and IP address, for security auditing.
申请页提交的内容会存下：邮箱、联系人称呼、电话，以及你选填的公司／团队、站点域名、备注；同时记录提交时的 IP 与浏览器 UA，仅用于防刷与取证。这些信息只用于审核你的申请、开通账号和就此事联系你，不用于营销推送，也不提供给第三方。审核结果会以邮件通知你——发信通过邮件服务商 Resend 送达，收件地址与邮件正文会经其服务器传输。 A submission on the request page stores: email, contact name, phone, plus the optional company/team, site domain and notes you provide; the submitting IP and browser user-agent are also recorded, solely for abuse prevention. This information is used only to review your request, provision an account and contact you about it — never for marketing, never shared with third parties. The outcome is emailed to you; delivery goes through the email provider Resend, so the recipient address and message body pass through its servers.
| 访客原始 IP 地址Visitors' raw IP addresses | IP 仅在服务器内存中用于解析出国家/省/市，解析完即丢弃，从不写入数据库。访客事件表中不存在 IP 字段 IP is used in server memory only to derive country/province/city, then discarded. It is never written to the database — the visitor event table has no IP column |
| 跨站追踪Cross-site tracking | 访客 ID 是第一方的，仅在安装了代码的那一个网站内有效。我们不会、也无法把同一个人在不同客户网站上的行为关联起来 The visitor ID is first-party and valid only within the single site that installed the snippet. We do not and cannot link the same person's activity across different customer sites |
| 第三方 CookieThird-party cookies | 完全不使用。访客侧只用 localStorage（见第五节） Not used at all. On the visitor side we use only localStorage (see section 5) |
| 设备指纹Device fingerprinting | 不采集 Canvas、字体、屏幕分辨率、硬件参数等主动探测得来的指纹信号，也不做跨站关联。我们确实记录浏览器自动发送的 User-Agent 字符串（见 2.1「环境」）——它用于识别设备类型、爬虫与 App 内嵌浏览器，本身是常规服务器日志字段，但我们不讳言它具备一定的区分度 No canvas, font, screen-resolution or hardware signals are actively probed, and no cross-site correlation is performed. We do record the User-Agent string the browser sends automatically (see 2.1 "Environment") — it is used to identify device type, bots and in-app browsers, and is a standard server-log field, though we will not pretend it carries no distinguishing power |
| 精确位置Precise location | 不调用浏览器定位接口。地理信息止于城市级，且由 IP 推断而来，本身就是粗略的 The browser geolocation API is never called. Geography stops at city level and is inferred from IP, which is coarse by nature |
| 页面内容与输入Page content and user input | 不录屏、不记录鼠标轨迹、不读取表单内容、不采集页面文本 No session replay, no mouse tracking, no form contents, no page text |
https://geo-analytics.info/v/collect，全程 HTTPS 加密
The script in the visitor's browser sends events in batches to https://geo-analytics.info/v/collect over HTTPS我们在访客浏览器的 localStorage 中写入三项，均为第一方，且可由访客随时自行清除（浏览器"清除网站数据"即可）： We write three first-party entries to the visitor's localStorage. Visitors can clear them at any time via the browser's "clear site data" function:
| 名称Key | 内容与用途Contents and purpose |
|---|---|
_vtag_vid |
随机匿名访客 ID，用于区分新老访客Random anonymous visitor ID, used to tell new from returning visitors |
_vtag_sess |
当前会话的 ID 与起止时间，用于切分会话Current session ID and timestamps, used for session segmentation |
_vtag_land |
首次到访的来源，用于首触归因The first-visit source, used for first-touch attribution |
拒绝同意（见第六节）时，以上三项一项都不会写入。 When consent is denied (section 6), none of these are written.
产品默认按"已同意"运行：访客首次打开页面时即写入匿名 ID，此行为发生在任何同意动作之前。这是一种"选择退出"模型。 The product runs in "granted" mode by default: the anonymous ID is written when a visitor first opens a page, before any consent action. This is an opt-out model.
网站经营者若面向需要事先取得同意的地区（如欧盟 GDPR/ePrivacy）的访客，必须自行改为选择进入模型：在脚本运行前声明拒绝，或将整段代码置于同意横幅之后按需加载。产品提供的接口为 geotag('consent','denied') 与 geotag('consent','granted')，具体用法见接入指南。
Website operators serving visitors in jurisdictions that require prior consent (e.g. GDPR/ePrivacy) must switch to an opt-in model themselves: declare denial before the script runs, or load the snippet only after the consent banner is accepted. The interfaces are geotag('consent','denied') and geotag('consent','granted'); see the install guide.
拒绝状态下产品仍会上报页面浏览等匿名计数，但访客 ID 记为 anon、不写入任何本地存储，因此无法区分新老访客。
Under denial the product still reports anonymous counts such as page views, but the visitor ID is recorded as anon and nothing is stored locally, so new and returning visitors cannot be distinguished.
访客也可以自行选择：清除浏览器网站数据即可重置匿名 ID；使用浏览器隐私模式则每次访问都是全新身份。 Visitors have their own options: clearing site data resets the anonymous ID, and private browsing makes every visit a fresh identity.
| 数据Data | 保留期限Retention |
|---|---|
| 访客原始事件Raw visitor events | 不超过 24 个月，到期后清理No more than 24 months, cleared on expiry |
| 聚合统计结果Aggregated statistics | 不含个人标识，长期保留Contains no identifiers; retained long term |
| 客户账号Customer accounts | 服务存续期间；账号注销后删除For the life of the service; deleted after account closure |
| 管理操作审计Admin audit log | 不超过 24 个月No more than 24 months |
如果你是某个网站的访客：由于我们持有的访客 ID 是匿名随机串、不与你的真实身份关联，我们无法仅凭姓名或邮箱定位到你的数据。可行的做法是： If you are a visitor to a site: because the visitor ID we hold is a random anonymous string not linked to your real identity, we cannot locate your data from a name or email alone. The practical options are:
_vtag_vid），我们可据此查询或删除对应记录
If you can supply your visitor ID (visible as _vtag_vid in your browser's developer tools under localStorage), we can look up or delete the matching records如果你是我们的客户：可通过控制台查阅名下站点的全部数据；如需导出或删除，请联系我们。 If you are our customer: you can view all data for your sites in the console. For export or deletion, contact us.
本产品是提供给网站经营者的分析工具，不面向未成年人提供服务，也不主动识别访客年龄。若网站经营者的服务面向未成年人，应自行评估并履行相应的额外义务。 This product is an analytics tool for website operators. It is not directed at minors and does not attempt to determine a visitor's age. Operators whose services are directed at minors must assess and meet the additional obligations that apply.
本政策更新时会修改页首的"最近更新"日期；涉及处理目的、范围或方式的重大变更，会另行通知客户。 When this policy changes we update the "last updated" date at the top. Material changes to the purpose, scope or means of processing will be communicated to customers separately.
| 运营主体Operating entity | 上海超响应数字科技有限公司 |
| 隐私事务联系邮箱Privacy contact | support@sr007.com |
| ICP 备案号ICP filing | 沪ICP备2023017581号-6 |