A well-curated dashboard helps both people and the AI Assistant quickly understand what matters, how performance is changing, and where to dig deeper.
This article focuses on how to curate the analysis in a dashboard: what questions it should answer, how queries should be structured, and how tiles should work together to support natural exploration. The goal here is to help you build dashboards that surface the right information clearly and make follow-up analysis easy for both users and the AI.
Think of the AI Assistant as another user sitting next to you. If the dashboard makes sense to a human, it will work well for the AI too.
Start With Clear Business Questions
Every strong dashboard starts with a small set of clear business questions. If you can’t easily describe what the dashboard is meant to answer, it will likely become unfocused and harder to use.
Ask yourself:
What decisions or reviews does this dashboard support?
What questions do I regularly ask when I open this dashboard?
What would I want to understand within the first 30 seconds?
Example: Pick-Up Dashboard
Instead of “show everything about bookings,” your questions might be:
How is pick-up pacing versus the same time last period?
Which future periods are accelerating or slowing?
Which segments are driving the change?
Are there specific dates that look unusually strong or weak?
Those questions guide which queries belong on the dashboard and which do not.
If a tile does not help answer one of the core questions, it probably does not belong on this dashboard.
Keep the Dashboard Focused and Intentional
A dashboard should support a specific type of analysis, not try to answer every possible question at once. Large dashboards with dozens of tiles often create more friction than clarity. Users end up scrolling, hunting for the right table, or recreating analysis elsewhere.
A focused dashboard:
Has a clear analytical purpose
Avoids duplicative or marginal tiles
Makes it obvious what deserves attention
This does not mean less data overall. It means curating the right data into the right tiles so the signal is clear.
If you find yourself adding tiles “just in case,” that is often a sign the dashboard is drifting away from its core purpose.
Follow a Natural Analytical Flow
Most analysis naturally moves from higher-level signals into deeper detail. Your dashboard should reflect that progression.
A common flow looks like:
High-level outcome or trend
What is happening overall?Drivers or contributors
What is influencing that change?Breakdowns or diagnostics
Where exactly is this coming from?
Example: Pick-Up Analysis
Tile 1: Total pick-up versus same time last period over time
Tile 2: Pick-up by segment to see which segments are driving the change
Tile 3: Pick-up by stay period or date to isolate specific peaks or valleys
A user can quickly orient themselves and then go deeper naturally. The AI Assistant can also use this layered context when answering follow-up questions.
A Single Tile Can Carry More Context Than It Shows
A tile does not need to display every measure it contains.
You might show a single KPI visually, but include additional measures in the query underneath that provide useful context:
A chart showing Revenue may also include Units and ADR
A chart showing Occupancy may also include Capacity and Units
A trend line may include prior period or pacing measures for comparison
Those additional measures can appear in tooltips for users who hover, and they are also available to the AI Assistant when answering questions.
This allows you to keep the dashboard visually focused while still preserving analytical depth.
Think in Terms of Repeatable Analysis
Dashboards work best when they support questions you and your team ask repeatedly.
Good candidates for dashboards:
Pick-up monitoring
Forecast versus budget tracking
Market or competitive benchmarking
If an analysis is highly exploratory or one-off, it may belong in a workbook instead. Dashboards should make common workflows faster and more consistent.
In some cases, dashboards also work together:
A high-level dashboard surfaces trends and opportunities
A more detailed dashboard supports deeper investigation
For example:
Upgrade Performance & Trends surfaces overall behavior
Upgrade Impact dives into financial and operational implications
Curate for Ease of Use, Not Volume
If users have to scroll extensively or search across many tiles to find what they need, the dashboard is likely doing too much.
Strong curation prioritizes:
Clarity over completeness
Signal over noise
Flow over density
Remember that the goal is not to reduce information, but to organize it so answers are easy to reach.
Why This Matters for the AI Assistant
The AI Assistant uses the dashboard as its starting context. When a dashboard is well curated:
The AI can understand what the analysis is about
It can connect insights across tiles more effectively
Follow-up questions become more accurate and relevant
Users get faster, clearer answers
If a dashboard is unfocused, overloaded, or inconsistent, both users and the AI will struggle to navigate it.
In practice, good dashboard curation improves the experience for everyone.
Final Thought
If a dashboard is easy for a person to understand, it will be easy for the AI to work with too.
Start with clear questions, keep the scope focused, structure analysis in a natural flow, and curate tiles intentionally. The AI Assistant then becomes a powerful extension of that work, helping you explore, validate, and learn faster.