[[["易于理解","easyToUnderstand","thumb-up"],["解决了我的问题","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["很难理解","hardToUnderstand","thumb-down"],["信息或示例代码不正确","incorrectInformationOrSampleCode","thumb-down"],["没有我需要的信息/示例","missingTheInformationSamplesINeed","thumb-down"],["翻译问题","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["最后更新时间 (UTC):2025-09-05。"],[],[],null,["# Best practices for rolling out Conversational Analytics with Looker\n\n\u003cbr /\u003e\n\n|\n| **Preview**\n|\n|\n| This product or feature is subject to the \"Pre-GA Offerings Terms\" in the General Service Terms section\n| of the [Service Specific Terms](/terms/service-terms#1).\n|\n| Pre-GA products and features are available \"as is\" and might have limited support.\n|\n| For more information, see the\n| [launch stage descriptions](/products#product-launch-stages).\n\n[Conversational Analytics](/looker/docs/studio/conversational-analytics) lets users query data that is modeled in LookML by asking natural language questions [within a Looker instance](/looker/docs/studio/conversational-analytics-looker).\n\nThis guide provides strategies and best practices to help Looker administrators and LookML developers successfully configure, deploy, and optimize Conversational Analytics. This guide covers the following topics:\n\n- [LookML best practices for Conversational Analytics](#lookml-best-practices-for-conversational-analytics)\n- [When to add context to LookML versus Conversational Analytics](#when-to-add-context-to-lookml-versus-conversational-analytics)\n- [Recommended setup and rollout strategy](#recommended-setup-and-rollout-strategy)\n\nBy preparing your LookML model and Conversational Analytics, you can increase user adoption and ensure that users get accurate and useful answers to their questions.\n\nLearn [how and when Gemini\nfor Google Cloud uses your data](/gemini/docs/discover/data-governance).\n| As an early-stage technology, Gemini for Google Cloud\n| products can generate output that seems plausible but is factually incorrect. We recommend that you\n| validate all output from Gemini for Google Cloud products before you use it.\n| For more information, see\n| [Gemini for Google Cloud and responsible AI](/gemini/docs/discover/responsible-ai).\n\nLookML best practices for Conversational Analytics\n--------------------------------------------------\n\nConversational Analytics interprets natural language questions by leveraging two primary inputs:\n\n1. **The LookML model**: Conversational Analytics analyzes the structure, fields (dimensions, measures), labels, and descriptions that are defined within the Looker Explores.\n\n2. **Distinct field values**: Conversational Analytics examines the data values within fields (specifically, string dimensions) to identify the available categories and entities that users might ask about. Cardinality (the number of unique values) can influence how these values are used.\n\nWhile powerful, the ability of Conversational Analytics to be effective is directly tied to the quality and clarity of these two inputs. The following table contains common ways that unclear or ambiguous LookML can negatively affect Conversational Analytics, along with solutions for improving the output and user experience.\n\nFor more best practices for writing clean, efficient LookML, see the following documentation:\n\n- [Best practice: What to do and what not to do with LookML](/looker/docs/best-practices/best-practices-lookml-dos-and-donts)\n- [Best practice: Create a positive experience for Looker users](/looker/docs/best-practices/how-to-create-a-positive-experience-for-looker-users)\n- [Best practice: Writing sustainable, maintainable LookML](/looker/docs/best-practices/how-to-write-sustainable-maintainable-lookml)\n\nWhen to add context to LookML versus Conversational Analytics\n-------------------------------------------------------------\n\nIn Conversational Analytics, you can add context inputs, such as field synonyms and [descriptions](/looker/docs/reference/param-field-description), both to LookML and inside [agent instructions](/looker/docs/studio/conversational-data-agents#write-agent-instructions). When you're deciding where to add context, apply the following guidance: Context that is always true should be added directly to your LookML model. Looker Explores may be used multiple places, including both in dashboards and in Conversational Analytics, so context that's applied in LookML must hold true for all possible users who will interact with the data.\n\nAgent context should be qualitative and focused on the user, and there can be many agents serving different users from one Explore. Examples of context that should be included in agent instructions, but not in LookML, are as follows:\n\n- Who is the user that is interacting with the agent? What is their role? Are they internal or external to the company? What is their previous analytics experience?\n- What is the goal of the user? What type of decision are they looking to make at the end of the conversation?\n- What are some types of questions that this user will ask?\n- What are the top fields that are specific to this user? What are fields that this user will never need to use?\n\nRecommended setup and rollout strategy\n--------------------------------------\n\nThis guide recommends the following phased approach for implementing Conversational Analytics in Looker:\n\n- [Phase 1: Curate data and define the initial scope](#phase-1)\n- [Phase 2: Configure agents and validate internally](#phase-2)\n- [Phase 3: Expand Conversational Analytics adoption to more users](#phase-3)\n\nThis approach lets you start with a small, controlled scope, validate your setup, and then expand to more users and data.\n\n### Phase 1: Curate data and define the initial scope\n\nIn this phase, prepare your data for users to query with Conversational Analytics and define the scope of the initial deployment. Follow these recommendations for starting with a small and controlled scope:\n\n- **Limit initial user access** : To enable internal testing and validation, use Looker's permission system to grant the [Gemini role](/looker/docs/admin-panel-users-roles#gemini) to a small set of users who are familiar with the data.\n- **Limit Looker model access for Gemini** : When you grant the [Gemini role](/looker/docs/admin-panel-users-roles#gemini), you can also limit which models Gemini can access. To start, consider limiting Gemini access to one or two models that you have curated for Conversational Analytics.\n- **Select curated Explores** : Start with one or two well-structured Explores that are based on relatively clean data and that provide clear business value. Optimize these Explores for Conversational Analytics in Looker by following the detailed instructions in [LookML best practices for Conversational Analytics](#lookml-best-practices-for-conversational-analytics).\n\n### Phase 2: Configure agents and validate internally\n\nIn this phase, build and refine your Conversational Analytics agents, and then thoroughly test them with internal users to confirm accuracy and effectiveness. This phase involves the following steps:\n\n1. **Create curated agents**: Create Conversational Analytics agents that are based only on the curated Explores that you prepared during the curation and initial setup phase.\n2. **Refine with agent instructions**: Use agent instructions to provide additional context and further guidance. For example:\n\n - Define synonyms for field names or values.\n - Provide specific context or rules for how certain fields should be used.\n\n | **Note:** Agent instructions can diverge from your LookML. Be mindful of potential discrepancies.\n3. **Validate internally and iterate**: Thoroughly test the agents with users who are familiar with the data. Ask various questions, test edge cases, and identify weaknesses. Make the following changes based on feedback from testing:\n\n 1. Refine the LookML. For example, adjust the values for the [`label`](/looker/docs/reference/param-field-label), [`description`](/looker/docs/reference/param-field-description), or [`hidden`](/looker/docs/reference/param-field-hidden) LookML parameters.\n 2. Adjust agent instructions.\n 3. Continue flagging issues with data quality.\n\n### Phase 3: Expand Conversational Analytics adoption to more users\n\nIn this phase, expand Conversational Analytics adoption to more users by granting access, collecting feedback, and iterating on your agents. This phase involves the following steps:\n\n1. **Grant targeted access** : Grant Conversational Analytics access to additional users who have the [Gemini role](/looker/docs/admin-panel-users-roles#gemini), and encourage those users to use the specific, vetted agents that you have created.\n2. **Launch and collect feedback**: Actively solicit feedback on the following topics:\n\n - Accuracy of responses\n - Ease of use\n - Missing information or confusing results\n3. **Iterate continuously**: Use feedback to make further refinements to LookML and agent instructions, and prioritize data cleanup efforts.\n\n4. **Expand access** : Once the agents prove stable and valuable, expand access to other relevant user groups and introduce new curated agents by granting the [Gemini role](/looker/docs/admin-panel-users-roles#gemini). You can also introduce new curated agents and expand access to the models that are available to the Gemini role, following the same processes that were used in the previous phases."]]