Skip to main content

Maximizing the Value of AI Summaries on Dashboards: Capabilities, Limitations, and Best Practices

The additional context window in Insights is a powerful tool that allows you to customize AI-generated chart summaries

Ashley Dehertogh avatar
Written by Ashley Dehertogh
Updated over 2 weeks ago

What Are AI Summaries?

AI Summaries as a Chart option in Insights provide automated, customizable interpretations of your data visualizations. These summaries help you quickly grasp key trends, anomalies, and insights without manually analyzing each detail on your results table.

If you're new to Insights' AI features, start with How Insights Uses AI in the Product for a complete overview.

What do AI Summaries Include?

A typical summary contains:

  1. Headline insights that highlight key takeaways

  2. Contextual explanations to clarify trends

  3. Notable patterns or anomalies that stand out

  4. Suggestions for further analysis or next steps

💡 Tip: Use AI summaries to accelerate your analysis—not replace it. Combine them with your expertise to draw meaningful conclusions.

How to Access AI Summaries?

  1. Navigate to the Chart in your workbook or dashboard

  2. Click the AI Summary chart icon

  3. The summary will display automatically

You can also tailor the output using the Additional Context input. This gives the AI clearer instructions on what you’d like to focus on.

Getting the Most Out of AI Summaries

1. Structuring Your Table for Better AI Summaries

The quality of your AI summary depends on how well your table is structured. Follow these best practices:

Essential Structure Tips

  • Enable column totals: This helps the AI understand overall context and proportions.

  • Add subtotals (when relevant): If you're analyzing hierarchical data (e.g., regions containing properties), add subtotals to provide the AI with important group-level context. The AI will use these subtotals to identify group trends and patterns.

  • Include the Currency Code dimension: Always add the Currency Code dimension when using revenue-related KPIs (ADR, Revenue, RevPAR, etc.,). This ensures the AI uses the correct currency symbol in its analysis and provides properly formatted monetary references.

  • Sort data meaningfully: Sort your data in a logical order (chronological for time periods, descending by value for performance metrics).

  • Limit excessive dimensions: Avoid overwhelming the AI; prioritize key groupings.

  • Include comparison metrics together: When analyzing period-over-period performance, include both current and comparison metrics (e.g., ADR and ADR STLP) together in your table for better comparative analysis.

Advanced Table Structure Tips

  • Group related metrics: Place related metrics next to each other (e.g., Occupancy, ADR, RevPAR) to help the AI understand their relationships.

  • Field ordering: Place dimensions first, followed by metrics in a logical grouping (Units → ADR → Revenue)

  • Row limits: Avoid excessive row limits that might truncate important data

  • Pivot configuration: When using pivoted data, ensure pivot fields are meaningful for comparative analysis

  • Calculated fields: Use calculated fields to derive important metrics not directly available in the topic

2. Customizing the Summary with Context Instructions

For more relevant insights, use the Additional Context field near the summary. A clear prompt helps guide the AI.

Sample prompt Structures:

  • Focus Areas: "Focus analysis on the weekday vs. weekend performance patterns visible in this chart"

  • Metrics Emphasis: "Emphasize RevPAR and its driving components shown in the visualization"

  • Comparisons: "Compare the current year to previous year patterns shown in the data"

  • Format Preferences: "Include clearly marked sections for overall performance, anomalies, and observations"

Example Context Templates

Here are ready-to-use templates you can copy and adapt:

For Performance Summaries

Create a detailed, analytical summary of the property's performance compared to the previous period. Highlight notable trends and any data anomalies. Conclude the summary with an ‘Actions to be Taken’ section clearly outlining areas needing further investigation. Use precise, professional language and support statements with numerical data.

For Pick-up Summaries

Summarize the property's pick-up performance clearly, highlighting key trends and anomalies. Use precise, professional language, focusing on actionable insights without suggesting improvements.

More context templates

Template 1: Performance Overview Analysis

Please provide a comprehensive analysis with these sections:
1. Overall Performance Summary (key headline about our performance)
2. Monthly Trends and Patterns (analyze seasonality and month-over-month changes)
3. Critical Metrics Analysis (focus on Occupancy, ADR, and RevPAR relationships)
4. Areas of Strength (identify top-performing periods and key factors)
5. Areas for Improvement (highlight underperforming periods and likely causes)
6. Forward-Looking Insights (what the current trends suggest for future periods)
Note: Our high season is May-August. Please focus on the relationship between occupancy and ADR, specifically which is driving our RevPAR performance.

Template 2: Period-Over-Period Comparison

Create a detailed period-over-period performance analysis with:
1. Headline Performance Finding (overall trend vs. previous period)
2. KPI Breakdown (analyze each KPI separately - Units, ADR, and Revenue)
3. Critical Variance Analysis (identify and explain largest variances)
4. Pattern Recognition (note any repeating patterns or trend shifts)
5. Performance Drivers (which KPIs are most influencing overall results)
Focus particularly on how our current performance compares to STLP and Previous, and whether we're seeing improvement in both occupancy and rate. Note that we increased rates by approximately 5% starting in March.

Template 3: Segment Analysis

Provide a segment analysis with the following structure:
1. Segment Distribution Overview (breakdown of business by segment)
2. High-Value Segment Performance (identify and analyze top revenue-generating segments)
3. Year-over-Year Segment Shifts (how has our segment mix changed)
4. ADR and Occupancy Analysis by Segment (pricing power and demand by segment)
5. Opportunity Identification (which segments show growth potential)
Context: Our strategic focus is on growing the Corporate and Group segments. We've implemented new corporate negotiated rates in February and launched a new group sales initiative in April.

Template 4: Channel Performance Review

Create a channel performance analysis structured as follows:Channel Mix Summary (overall distribution of business by channel)Direct vs. OTA Performance (compare direct booking performance to OTA channels)Profitability Analysis (examine net contribution after accounting for costs)Year-over-Year Channel Shift (identify meaningful changes in distribution)Strategic Recommendations (based on the data, what channel strategy adjustments should we consider)Note that we increased our marketing spend on direct booking initiatives in March, and OTA commission costs have increased by 2% since January. Our goal is to increase direct bookings by 10% this year.

Template 5: Geographic Source Market Analysis

Analyze our geographic source markets with these components:Market Distribution Overview (where our business is coming from)High-Value Markets (identify top markets by total revenue and ADR)Emerging Markets (markets showing strongest year-over-year growth)Seasonal Market Patterns (how source market mix shifts throughout the year)Strategic Opportunities (which markets warrant more marketing attention)Context: We've increased marketing efforts in the Northeast region and reduced European marketing. Our airport now has new direct flights from Chicago and Denver that started in April.

Template 6: Day of Week Performance

Structure a day-of-week performance analysis with:Day of Week Distribution (overall business patterns by day)Weekday vs. Weekend Performance (compare business patterns and metrics)Day-Specific Opportunities (identify specific days with improvement potential)Rate and Occupancy Strategy (analyze pricing power by day of week)Recommended Focus Areas (which days warrant strategy adjustments)Note that our property primarily serves business travelers Monday-Thursday and leisure on weekends. Our restaurant is closed on Mondays, and we've recently introduced a Sunday package that includes late checkout.

Template 7: Lead Time Analysis

Provide a detailed booking lead time analysis:Lead Time Overview (summarize booking window patterns)Segment-Specific Patterns (how lead time varies by business segment)ADR Correlation (relationship between booking timing and achieved rates)Year-over-Year Comparison (how current lead times compare to previous periods)Strategic Implications (what these patterns mean for pricing and revenue strategy)Context: Our cancellation policy changes at 14 days prior to arrival. We implement advance purchase discounts for bookings 45+ days out, and last-minute rates within 3 days of arrival.

Template 8: Executive Summary

Create an executive-level summary suitable for senior leadership:Headline Performance (one strong, data-driven statement about overall performance)Key Business Metrics (concise summary of critical KPIs)Notable Successes (1-2 specific areas of strong performance)Watch Areas (1-2 specific areas requiring attention)Strategic Implications (what this data suggests for our strategy)Use business language rather than technical metrics where possible, and focus on bottom-line impact. This summary will be shared with our ownership group who are primarily interested in RevPAR performance versus budget and previous year.

Template 9: Detailed Operational Analysis

Provide a comprehensive operational analysis for our management team:Performance Overview (overall performance summary)Occupancy Analysis (detailed breakdown of occupancy patterns)Rate Strategy Assessment (evaluation of our pricing strategy effectiveness)Revenue Management Opportunities (specific opportunities to optimize revenue)Operational Considerations (implications for staffing and service levels)Include specific data points and percentages where relevant. This analysis will be used by our operations team to plan staffing and resource allocation, so please highlight day-of-week and monthly patterns clearly.

Template 10: Forward-Looking Performance Assessment

Create a forward-looking analysis based on the on-the-books data:Booking Pace Assessment (are we ahead or behind pace)Rate Integrity Evaluation (are we maintaining ADR while building occupancy)Critical Period Identification (highlight periods needing attention)Segment and Channel Mix (how our business mix is shaping for future periods)Strategic Recommendations (suggested actions based on current pace)Context: We typically need 70% occupancy to break even. Our key upcoming event periods are July 15-25 and October 8-15. We've implemented new pricing strategies in August.

Customizing Templates for Your Needs

Tailor templates by including:

  1. Add property-specific context:

    • Mention recent renovations, service changes, or market developments

    • Include information about local events or demand drivers

    • Reference specific business goals or KPI targets

  2. Specify time periods of interest:

    • Highlight specific months, seasons, or days requiring special attention

    • Reference important historical periods for comparison

    • Identify upcoming high-demand periods

  3. Include industry-specific terminology:

    • Use terms familiar to your team and stakeholders

    • Reference relevant benchmarks or industry standards

    • Include property-specific segmentation or categorization

  4. Tailor to your audience:

    • Adjust language and focus for operational vs. executive audiences

    • Emphasize financial metrics for ownership groups

    • Focus on operational metrics for management teams

Understanding Limitations

AI Summaries are powerful, but not perfect. Keep these limitations in mind:

  • Data Limitations

    • AI only analyzes what’s visible in your Results Table

    • It can’t pull data from other charts, dashboards, or prior queries

  • Context Blind Spots

    • AI doesn’t know your goals, business model, or market unless you tell it

    • You must provide details like events, promotions, or property changes

  • Technical & Analytical Constraints

    • It performs best with clear patterns, not subtle signals

    • Summaries may simplify complex logic or nuanced segmentation

    • Cannot infer or simulate strategy-specific logic (e.g., discount tiers)

Best Practices for Great Results

  1. Start with properly structured data: Ensure your visualization has appropriate totals, currency codes, and comparison metrics

  2. Be specific in your context instructions: Direct the AI's attention to particular aspects of your data

  3. Provide business context not visible in the data: Add information about strategy changes, market conditions, or property-specific factors

  4. Request a logical structure: Ask for clear sections that build on each other

  5. Iterate your instructions: If the initial summary doesn't meet your needs, refine your context instructions

  6. Combine multiple templates: Mix elements from different templates to create customized instructions

  7. Specify format preferences: Request bullet points, narrative paragraphs, or another format that suits your needs

Troubleshooting Common Issues

If...

Then try...

Example Prompt to Add in Context Window

The summary is too general

Add more specific metrics or thresholds to focus the analysis

“Highlight any days where ADR exceeds $250 and occupancy is above 85%.”

Key trends are missing

Explicitly call out the trends or comparisons you want the AI to analyze

“Focus on how ADR and occupancy have changed compared to the previous month.”

You need a clear comparison

Name the comparison periods (e.g., STLP, Previous, YoY) for accurate analysis

“Compare performance this month to STLP and Previous actuals.”

You want actionable insights

Ask the AI to suggest next steps or strategy recommendations

“Include a section with recommended actions based on underperformance.”

The language is too technical

Request simpler, business-friendly phrasing for a broader audience

“Summarize results in plain language that’s accessible to non-revenue managers.”

You need a structured format

Direct the AI to organize insights clearly using labeled sections

“Use this structure: 1. Overall Performance,
​2. Key Trends,
​3. Suggested Actions.”

The summary skips context

Provide your own details about events or strategy changes

“Note that we increased rates by 5% in March and launched a new package in April.”

Insights feel disconnected

Ask the AI to link KPIs (e.g., how ADR and occupancy affect RevPAR)

“Analyze how occupancy and ADR together are influencing RevPAR performance.”

Final Takeaway

AI Summaries enhance your ability to interpret visual data—fast. But to unlock their full potential, structure your results well, guide them with context, and combine the AI’s observations with your expertise. The best outcomes come from collaboration between data, AI, and human insight.


📌 Need more information? Reach out to our Advisory team via the chat with any feedback on particularly good examples you've used and/or you need any support!

Did this answer your question?