Skip to main content
Data Prompts

Dashboard Specification Writer

Generate a complete specification for a data dashboard — metrics, charts, filters, and layout — ready to hand to a designer or BI developer.

intermediateWorks with any modelData
Prompt
Write a complete specification for a data dashboard for [AUDIENCE] to monitor [BUSINESS_GOAL].

**Data sources available:** [DATA_SOURCES]
(e.g., "PostgreSQL sales DB, Stripe payments API, Google Analytics 4")

**Key questions the dashboard must answer:**
[KEY_QUESTIONS]
(List 3–6 specific business questions the dashboard should answer — e.g., "Which sales rep closed the most deals last month?", "Where are users dropping off in the checkout funnel?")

**Update frequency:** [REFRESH_RATE]
(e.g., real-time, hourly, daily, weekly)

Produce a dashboard specification that includes:

1. **Dashboard title and purpose** — one-sentence description of who uses it and when.

2. **KPI summary row** — 4–6 headline metrics shown as single numbers at the top (metric name, definition, data source, and target or benchmark if applicable).

3. **Chart specifications** — for each of the [KEY_QUESTIONS], specify:
   - Chart type and why it best answers the question
   - X-axis, Y-axis, series/grouping
   - Filters and drill-down options
   - Time range controls

4. **Filters panel** — global filters that apply to the entire dashboard (date range, segment, region, etc.)

5. **Layout recommendation** — section order and rough grid layout (e.g., "2-column grid: KPIs top row, funnel left, trend line right").

6. **Access and refresh** — who should see this dashboard and how often data should refresh.

How to Use

Describe the people who'll use the dashboard in [AUDIENCE] (e.g., "VP of Sales and regional sales managers"). Set [BUSINESS_GOAL] to the monitoring objective (e.g., "track revenue pipeline health"). List the specific questions the dashboard must answer — vague questions produce generic specs; specific ones produce actionable ones.

Variables

VariableDescription
[AUDIENCE]Who will use the dashboard and how often
[BUSINESS_GOAL]What the dashboard is for — the monitoring or decision-making objective
[DATA_SOURCES]Systems and databases where the data lives
[KEY_QUESTIONS]3–6 specific business questions the dashboard must answer
[REFRESH_RATE]How frequently data should update

Tips

  • Start with the questions, not the charts — good dashboards answer specific questions; bad ones are a collection of charts in search of a question.
  • If you already have a rough sketch, describe it in the prompt and ask the AI to critique it instead: "Identify any missing metrics or questions that this layout doesn't answer."
  • After generating the spec, paste it to a BI developer or into a tool like Retool, Metabase, or Looker and ask it to generate the underlying SQL queries for each chart.