我们将于 2024 年 7 月 31 日逐步淘汰 Google Business Messages。点击
此处了解详情。
设计前提条件
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。
您提供的对话类型取决于客户的需求和您的业务目标。为了准备对话设计,请先收集以下信息。
- 您的品牌风格如何?聊天机器人代表的是您的品牌,因此它应该采用您品牌的语气。记住这一点有助于您编写聊天机器人的回复。
- 对话的目的是什么?本地商店信息、客户服务、日程安排、辅助购物使用此意图定义代理的核心功能。
- 您的代理支持哪些入口点?换句话说,您的用户来自哪里?
尝试根据入口点预测常见的用户意图。
- 从 Google 地图开始对话的用户可能需要特定于商店的信息,如商品目录或营业时间。
- 从非本地入口点(如 Google 搜索)开始,用户可能需要更宽泛的信息,如订单状态或退货政策。
- 从品牌网页开始的用户可能对自己的帐号有疑问。
- 用户的目标是什么?换句话说,您的用户想要去哪里?了解转化历程的起点和终点有助于您设计从头到尾的理想路径。
- 聊天机器人可以在历程的哪些方面提供服务?确定用户可以使用自动化功能和自助客服的位置。这有助于您定义代理的功能并规划适当的切换响应。
- 代理的发布计划是什么?某些用例可能是季节性的,或者与促销活动有关。了解发布计划有助于您针对更有可能是特定日期或事件的用户查询做好准备。
- 出色对话涉及哪些指标?了解成功的标准提高了客户服务工作效率、潜在客户转化次数、解决与商店相关的问题、客户满意度较高的得分?这些指标可帮助您评估和优化对话体验。
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):2023-01-03。
[null,null,["最后更新时间 (UTC):2023-01-03。"],[[["\u003cp\u003eDefine your brand's tone of voice and the chatbot's purpose, such as customer support or scheduling, before designing conversations.\u003c/p\u003e\n"],["\u003cp\u003eAnticipate user intents based on entry points like Google Maps, Search, or your webpage to tailor responses and functionalities.\u003c/p\u003e\n"],["\u003cp\u003eMap the user journey to identify areas where a chatbot can effectively automate tasks versus where human intervention is necessary.\u003c/p\u003e\n"],["\u003cp\u003eConsider the rollout plan and potential seasonal or promotional influences to prepare for specific user queries.\u003c/p\u003e\n"],["\u003cp\u003eEstablish clear metrics, like customer satisfaction or lead conversions, to measure the success and effectiveness of your chatbot interactions.\u003c/p\u003e\n"]]],[],null,["# Design prerequisites\n\nThe type of conversation you offer depends on your customers' needs and your\nbusiness objectives. To prepare for conversation design, gather the following\ninformation first.\n\n- **What is the tone of your brand?** The chatbot represents your brand, so it should use your brand's tone of voice. Keeping this tone in mind helps you write the bot's responses.\n- **What is the purpose of the conversation?** Local store information, customer support, scheduling, assisted shopping? Use this intent to define the agent's core functionality.\n- **Which [entry points](/business-communications/business-messages/guides/concepts/entry-point)\n does your agent support?** In other words, where are your users coming from? Try to anticipate common user intents based on entry point.\n - Users who start a conversation from Google Maps may want store-specific information like inventory or opening hours.\n - Users who start from a non-local entry point like Google Search may want broader information like order status or return policy.\n - Users who start from a brand's webpage may have questions about their account.\n- **What are the users' goals?** In other words, where do your users want to go? Knowing where their journey begins and ends helps you design the ideal path leading from start to finish.\n- **Which aspects of the journey can a chatbot serve?** Identify where users can self-serve with automation vs. where a live agent is needed. This helps you define the agent's capabilities and plan an appropriate handoff response.\n- **What is the rollout plan for the agent?** Some use cases may be seasonal or tied to promotional events. Knowing the rollout plan helps you prepare for user queries that are more likely around specific dates or events.\n- **What are your metrics for a good conversation?** Know what success looks like. Increased customer service productivity, lead conversions, resolution of store-related questions, high CSAT scores? These metrics help you assess and refine the conversational experience."]]