Introduction
To ensure effective usage and maximize the benefits of Docket, it is crucial for customers to understand how to best leverage its capabilities and what is required for optimal performance. This guide will help you connect the right information sources, add essential content, and ask questions in ways that yield the best results.
Connecting Information Sources
To enhance Docket's ability to deliver accurate and relevant answers, integrate various information sources from day one. The more diverse and extensive the sources, the better the coverage of answers. Recommended source types include:
Sales Content Management: Examples include HighSpot, Seismic, Google Drive.
Communication: Utilize tools like Slack.
Internal Articles: Examples include Intercom, Notion.
Call Recordings: Tools such as Chorus, Gong.
Connecting these sources ensures Docket has access to a wide range of data, improving its ability to provide accurate and relevant answers.
Adding Essential Content
For Docket to function optimally, it should be equipped with key content types. These include:
Content Type | Description | Utilization |
Battlecards and Competitive Analysis | Detailed competitor comparisons. | Ingest battlecards for accurate product comparisons. |
Product FAQs and Documentation | FAQs and documentation. | Ingest FAQs to answer common feature and troubleshooting queries. |
Marketing Collaterals | Brochures, feature sheets, pricing, security info. | Ingest marketing materials for detailed product insights. |
External Calls from Stellar Account Executives | Call recordings of high-performing solution engineers' and account executives' calls. | Ingest these calls to increase knowledge coverage for Docket to learn about product updates, launch updates, and pricing packages. |
Leverage Communication Channels | Real-time interactions and updates. | Add Docket to channels where product information is shared and where Account Executives (AEs) ask questions to Subject Matter Experts (SMEs). This allows Docket to learn from real-time interactions and updates, enhancing its knowledge base dynamically. |
Questions Docket Can Answer Effectively
Question Type | Description | Example | Effectiveness |
Direct Questions | Requires a straightforward, specific answer about the company, its products, or services. | What is Slack Canvas? | Docket excels at answering direct questions using both structured sources like FAQs and documents, and unstructured sources like Slack messages or calls. |
Comparative Questions | Asks for a comparison between two or more items, features, or concepts related to the company and its competitors. | How does Slack compare to Microsoft Teams in terms of ease of use and features? | Docket can generate effective answers for comparative questions using structured sources like battlecards and documents. It can also provide satisfactory answers from unstructured sources if explicit facts are identified. |
Compound Questions | Contains multiple components that need to be addressed separately, often involving more complex or layered information. | What are the key features of Slack’s Enterprise Grid, and how does it differ from the standard version? | Docket can handle simple compounding questions, such as those about a product and its features. For questions involving multiple unrelated components that require a combination of different information sources, the answer quality might not be optimal. |
How-To Questions | Involves steps or processes that need to be followed. | What are the steps to integrate Slack with Google Drive? | Docket is effective at answering procedural questions if structured information is available within the company. However, it may not perform well with troubleshooting questions that require combining multiple sources or dealing with specific user issues not covered in common FAQs. |
Clarification Questions | Seeks to understand or clarify previous information. | You mentioned Slack Canvas in your previous answer. Can you explain how it can be used for team collaboration? | Docket is highly effective at answering clarification questions, particularly when the questions are short and concise. It can utilize unstructured sources effectively since only few references are generally used. |
Questions Docket Might Not Answer Optimally
Question Type | Description | Example | Effectiveness |
Hypothetical Questions | Poses a scenario to understand potential outcomes or procedures. | If a team needs to manage multiple projects simultaneously, how can Slack help streamline their workflow? | Docket may struggle with such questions, as they require reasoning based on knowledge of different outcomes or procedures not readily available in its knowledge base. |
Resource Availability and Location Questions | Questions asking for the availability and location of specific resources. | Do we already have product training resources for new hires that I can easily link to our New Hire Page? Where are the Loom product demo/training videos stored? | Docket currently provides answers to queries and includes references. However, it is not well-suited for handling search-intended questions about the availability and location of resources. Soon, Docket will support the file search intent as it's in the roadmap. |
Conclusion
For customers to effectively use Docket, it is crucial to integrate diverse and comprehensive information sources, incorporate essential content types, and leverage communication channels for continuous updates. By understanding the types of questions Docket can answer effectively and recognizing its limitations, users can set realistic expectations and optimize their experience with Docket.