Constructing Questions for Ask Zia

Ask Zia simplifies data interpretation by enabling users to analyze data using natural language. It eliminates the need to build complex reports manually or configure multiple groupings. Users can ask questions in natural language to analyze their data, ranging from descriptive analysis to predictive analysis. It can also highlight key drivers and explain why certain trends or results occurred, helping users quickly move from data to understanding.

• Who can Use Ask Zia - Access and Permissions
• Context Retention
• Level of Detail in Queries
• Best Practices for Using Ask Zia
• Descriptive Analysis
• Diagnostic Analysis
• Predictive Analysis
• Keywords for Ask Zia
   

Who can Use Ask Zia - Access and Permissions

Users with Administrator privileges, Shared users and Custom roles users, can interact with Ask Zia and perform conversational analysis on their data, subject to the data access permissions defined for their role.

Context Retention

Ask Zia maintains context throughout a conversation, understanding each query in relation to previous questions. This allows users to ask follow-up questions naturally without repeating the metrics or filters while continuing the conversation.

• Real-time type-ahead assistance suggests relevant fields such as columns, dimensions, measures, and functions, and may also recommend suitable visualizations or existing reports related to the query.
• Context-aware follow-up suggestions: During a conversation, Ask Zia analyzes the current query and proposes logical next steps, such as additional metrics, filters, or related analytical questions to refine the analysis.
• Proactive query recommendations: As users begin interacting with Ask Zia, it suggests relevant analytical questions by analyzing the data model, previous user interactions, and commonly used query patterns.

Level of Detail in Queries

The level of detail in a query determines how Ask Zia interprets the request and generates the resulting visualization. A query includes components such as metrics, dimensions, filters, aggregations, and geographic attributes. The combination of these elements defines the scope and granularity of the analysis.

By progressively adding these elements, users can move from a high-level overview to more focused and detailed insights. This flexible query structure allows Ask Zia to support both quick summaries and deeper exploratory analysis through natural language queries.

Metrics and Dimensions

Metrics represent the quantitative values being analyzed, such as revenue, sales, profit, or count of records. Dimensions provide context by grouping these metrics across categories like product, region, time period, or salesperson.

Combining metrics with one or more dimensions enables users to compare performance across categories, identify patterns, and analyze variations within the data.

Filters and Conditions

Filters allow users to restrict the dataset to specific criteria or time periods. Queries can include filters based on text values, numeric ranges, exact dates, date ranges, or relative time periods.

Applying filters helps narrow the analysis to relevant subsets of data, making it easier to focus on specific segments or conditions.

Geographic Analysis

Geographic attributes such as country, state, region, or city can be included in queries to analyze data by location. When geographic fields are used, Ask Zia automatically generates map-based visualizations to display regional distribution and spatial trends in the data.

Aggregations and Summary Functions

Aggregation enables users to summarize data using functions such as sum, average, count, minimum, and maximum. These functions help compute summary statistics and evaluate overall performance of a metric.

Ask Zia automatically interprets aggregation terms expressed in natural language and applies the appropriate calculation to the selected metric, allowing users to analyze summarized results without manually defining calculations.

Best Practices for Using Ask Zia

• Start with simple queries: Begin with a high-level question and progressively refine it with follow-up queries to explore the data further.
• Use clear metric and dimension names: Use terms that closely match the column names in your dataset so Ask Zia can interpret the query accurately.
• Include relevant filters: Apply filters such as time periods, categories, or conditions to narrow the analysis to the most relevant data.
• Group metrics by dimensions: Analyze metrics across one or more dimensions to compare performance across categories or segments.
• Leverage follow-up queries: Use conversational context to refine the analysis without repeating the entire question.
• Specify time periods for trend analysis: Include time context such as months, quarters, or years when analyzing trends or performance over time.
• Gradually increase query detail: Add dimensions, filters, or sorting progressively to move from high-level summaries to deeper insights.

Descriptive Analysis

Descriptive analysis focuses on understanding What has happened with the data. This includes summarizing metrics, comparing and identifying patterns. This is used to jump start data exploration.

Overall Performance

Provides a high-level summary of key business metrics such as total sales, revenue, or number of orders. It helps quickly understand the overall performance of the business during a specific period.

What is the total sales for Jan 2026

Category or Segment-wise Analysis

Analyzes performance across different segments such as region, product category, or customer group. This helps identify which segments contribute the most to sales.

Show me the total sales by Region and Product Category

Rank and Top Performers

Identifies entities with the highest or lowest values for a given metric, allowing to understand relative performance across different categories.

top 10 Product by Sales

Time-based Analysis

Examines how metrics change across different time periods such as days, months, or quarters. This helps track performance at specific points in time.

 Show monthly revenue for the last year

Correlation between Metrics

Analyzes whether two metrics move together or influence each other. For example, understanding how marketing spend relates to sales performance.

Correlation helps find if and how strongly two measures are related to each other. If a measure or variable increases or decreases together, they are positively correlated. If one measure tends to increase and another one decreases, then they are negatively correlated.

Zia computes the correlation using Pearson's Correlation Coefficient, which is the covariance of the two variables to the product of their standard deviations.

 What is the correlation between Customer Acquisition Cost and Revenue?

Trend Strength in the data

Measures the strength or degree of relationship between variables in the dataset. This helps determine how strongly one factor is associated with another.

Trend strength is a statistical measure used to determine how strongly data is moving in an upward or downward direction. It helps assess whether a trend is likely to continue or weaken.

Comparison Analysis

Compares performance across categories, regions, products, or time periods. This helps identify differences and determine which entities perform better.

 Compare sales across product categories for the last two years.

Breakdown Analysis

Breaks down a metric into multiple hierarchical levels for deeper insights. For example, analyzing sales by region, then by category, and then by product.

Time Series Analysis

Studies patterns and variations in data over continuous time intervals. It helps identify trends, seasonal patterns, and changes in performance over time.

 Show the revenue trend for the last two years

Diagnostic Analysis

Diagnostic Analysis focuses on understanding why something happened in the data. It goes beyond simply describing results and probes the root causes, relationships, and contributing factors behind an outcome.

Diagnostic analysis can be performed in different ways depending on what you want to investigate. It typically focuses on identifying the factors responsible for a change or pattern in the data. Diagnosis can be based on:

• Increase or decrease analysis: Examining why a metric such as sales, revenue, or user activity has increased or decreased over a specific period.
• Segment-based diagnosis: Analyzing performance across different segments such as region, product category, customer group, or sales representative to identify which segment contributed to the change.
• Behavior-based diagnosis: Understanding patterns in user or customer behavior, such as purchasing trends, churn patterns, or engagement changes, to determine what influenced the outcome.

Predictive Analysis

Predictive analysis focuses on estimating future outcomes and identifying unusual patterns based on historical data. By analyzing past trends and patterns, Ask Zia can generate forecasts and detect anomalies that help users anticipate changes in performance and identify unexpected behavior in their data.

Forecast

You can forecast the data by simply asking for the future data. The following example will predict how the future sales for the next 6 months that will most likely be.

Get me the sales for next 6 months

Trend Line

Trend analysis helps understand how a metric changes over time. By evaluating historical values across a selected time dimension, Ask Zia identifies patterns such as growth, decline, or seasonal fluctuations, enabling users to track the overall direction of performance.

Show me the Sales Trend

Forecasting Trend

Ask Zia can add both forecasting and trendline to you chart. The following question forecast the sales for the next 6 months and also predict the sales trend.

Show me the Sales Trend for next 6 months

Keywords for Ask Zia 

The following is the list of keywords that Ask Zia can interpret and create appropriate insights.