The Response Generation dashboard provides deep visibility into how your team uses SiftHub’s AI to answer questions and populate documents. These insights are primarily driven by Answer and Autofill activities happening across the SiftHub Browser Extension, Web App, and MS Office Add-in.
Response Generation insights currently exclude activities from the AI Teammate module.
Managing the Insights View
Before analyzing the data, use the dashboard controls to isolate the specific information you need:
Date Range Selector: View data for a specific timeframe (defaults to 30 days) to identify long-term adoption or response trends.
User & Tool Filtering: Click the Filter icon to drill down by specific Users or where the activity was Initiated from (e.g., specific bots or modules).
Export Data: Click the Download icon on charts (like the Question Volume Trend) to export raw data for external reporting.
Interactive Drill-down: Click on any specific graph item in the Question initiated from or Answer completion charts to automatically open their corresponding logs in a new tab with that filter already applied.
Overview Section
The Overview section provides an immediate snapshot of your organization's AI volume, accuracy, and efficiency.
Key Performance Indicators (KPIs)
The top row highlights cumulative metrics, each featuring a trend indicator (e.g., 30% ↘) that compares the current data to the previous equal time period:
Total questions asked: The cumulative number of individual queries processed during the selected period.
Total questions answered: The specific count of queries that received a successfully generated AI response.
Documents autofilled: The total number of unique questionnaires or proposals processed via bulk Autofill.
Average response time: The median time (in seconds) the AI takes to generate each individual answer.
Distribution Charts
Visual breakdowns help you understand tool adoption and output quality:
Question initiated from: A donut chart showing the split between entry points, such as Autofill, the Answer tab (Extension/MS Add-in), Slack, and Google Chat.
Answer completion: A breakdown of response quality, tracking the percentage of queries that were successfully Answered, resulted in a Partial response, or had No information found in your connected knowledge sources.
Question Volume Trend
This graph tracks the daily, weekly, or monthly volume of Total questions asked versus Questions answered. It is an essential tool for identifying peak activity periods, such as heavy RFP submission weeks or end-of-quarter surges.
Frequently Asked Questions (FAQ) Insights
The FAQ section is one of the most actionable parts of the Insights module, providing a semantic map of the inquiries your team receives from prospects and within questionnaires. It moves beyond simple lists to show you exactly where your knowledge base is strongest and where critical gaps exist.
The dashboard visualizes this data on two distinct fronts:
1. Frequently Asked Question Categories & Sub-categories
On the left side of the FAQ card, SiftHub automatically categorizes every question into logical buckets such as Product Capabilities, Security & Compliance, and Technical & Infrastructure:
Distribution and Trend: View what percentage of total questions fall into each category, accompanied by a trend indicator comparing interest to the previous period.
Drill-down into Sub-categories: Click "View sub-categories" on any main category to see a more granular breakdown (e.g., drilling into Security & Compliance to see Information Security Policy or Data Security & Encryption).
Direct Log Access: Clicking on any category or sub-category name immediately opens a new tab with the Usage Logs filtered to that specific group, allowing you to review every related question and answer.
2. Top Asked Questions (Semantic Grouping)
On the right side, SiftHub identifies "master questions". Instead of listing every unique variation, the AI uses semantic similarity to group similar queries together:
Semantic Grouping Example: Queries like "Do you have SSO?" and "Is SSO support available?" are grouped under one master question to show the true frequency of that specific requirement.
Frequency Tracking: See exactly how many times a specific semantic group has been triggered across all your sales activities.
Full List Review: Click "View all" to see the complete list of semantic question groups ranked by frequency.
Identifying Knowledge Gaps
You can toggle the entire FAQ view between All, Answered, and Unanswered questions. This is a critical tool for knowledge management:
All: A complete overview of every semantic group and category triggered.
Answered: Highlights high-frequency questions that the AI is successfully handling using your existing content.
Unanswered (Knowledge Gaps): This view isolates questions that the AI could not answer. By identifying which Categories and Semantic Groups are frequently unanswered, you can pinpoint exactly which documents or Q&As need to be added to the Repository to improve your team's automation rates.
Top Used Instructions Insights
Located at the bottom of the dashboard, this section identifies the most popular Instruction Prompts applied by your team to customize AI responses:
Prompt Strategy: Review which instructions (e.g., "Respond with a Yes or No only" or "Give a brief answer in bullet points") are being used most often.
Usage Frequency: Track how many times each specific instruction has been applied to generated answers during the selected period.
Optimization: Use this list to identify the most effective prompt strategies and share them with the wider team to standardize high-quality output.