Dashboard Purpose

Technical Explanation

A dashboard is a visual tool that displays key metrics and KPIs to monitor performance and enable data-driven decisions. Its primary purpose is to answer business questions at a glance, not to explore data in depth.

Dashboard vs Analysis

Aspect	Dashboard	Ad-hoc Analysis
Purpose	Monitor, track	Investigate, understand
Frequency	Daily/ongoing	When needed
Questions	Known in advance	Unknown beforehand
Audience	Broad	Specific

The Purpose Framework

Before building a dashboard, ask:

1. Who is viewing it?
2. What decision should this inform?
3. When do they need it?
4. Why does this metric matter?

If you can't answer these questions, you don't need a dashboard yet.

---

Code Examples

Using the CatCafe dataset:

-- Example: Executive Dashboard Queries

-- 1. Top-level KPIs (what executives want daily)
SELECT
    (
        SELECT COUNT(DISTINCT customer_id)
        FROM orders
        WHERE order_date >= CURRENT_DATE - INTERVAL '30 days'
    ) as active_customers_30d,

    (
        SELECT SUM(total_amount)
        FROM orders
        WHERE status = 'completed'
        AND order_date >= CURRENT_DATE - INTERVAL '30 days'
    ) as revenue_30d,

    (
        SELECT AVG(total_amount)
        FROM orders
        WHERE status = 'completed'
        AND order_date >= CURRENT_DATE - INTERVAL '30 days'
    ) as avg_order_value_30d,

    (
        SELECT COUNT(*)
        FROM customers
        WHERE registration_date >= CURRENT_DATE - INTERVAL '30 days'
    ) as new_customers_30d;

-- 2. Trend KPIs (month-over-month)
WITH monthly_revenue AS (
    SELECT
        DATE_TRUNC('month', order_date) as month,
        SUM(total_amount) as revenue
    FROM orders
    WHERE status = 'completed'
    GROUP BY DATE_TRUNC('month', order_date)
)
SELECT
    month,
    revenue,
    LAG(revenue) OVER (ORDER BY month) as prev_month,
    revenue - LAG(revenue) OVER (ORDER BY month) as mom_change,
    (revenue - LAG(revenue) OVER (ORDER BY month)) * 100.0
        / NULLIF(LAG(revenue) OVER (ORDER BY month), 0) as mom_change_pct
FROM monthly_revenue
ORDER BY month DESC;

-- 3. Funnel KPIs (conversion tracking)
WITH funnel AS (
    SELECT
        COUNT(DISTINCT user_id) FILTER (WHERE event_type = 'page_view') as visitors,
        COUNT(DISTINCT user_id) FILTER (WHERE event_type = 'add_to_cart') as cart_adds,
        COUNT(DISTINCT user_id) FILTER (WHERE event_type = 'purchase') as purchases
    FROM customer_events
    WHERE event_date >= CURRENT_DATE - INTERVAL '30 days'
)
SELECT
    visitors,
    cart_adds,
    purchases,
    cart_adds * 100.0 / NULLIF(visitors, 0) as view_to_cart_pct,
    purchases * 100.0 / NULLIF(cart_adds, 0) as cart_to_purchase_pct
FROM funnel;

---

The Cat Analogy

A dashboard is like a cat's health tracker collar:

What it shows:

• Current energy level (green/yellow/red)
• Steps taken today vs goal
• Water intake
• Sleep quality

What it's NOT:

• Detailed activity log (that's analysis)
• Historical data from 5 years ago (not relevant)
• Raw GPS coordinates (too granular)

The collar answers one question: "Is my cat healthy right now? Do I need to do something?"

That's the purpose: actionable status at a glance.

---

Common Pitfalls

Building Dashboards Before Knowing the Purpose

-- WRONG: "Let's put all our metrics on a dashboard"
SELECT * FROM all_our_tables;
-- This creates noise, not clarity

-- RIGHT: Start with the question
-- "What decision does this dashboard inform?"
-- If you can't answer, don't build it yet

Vanity Over Actionability

-- WRONG: Showing impressive numbers that don't drive action
SELECT
    total_users,          -- Vanity: doesn't tell you what to do
    total_page_views,      -- Vanity: high numbers feel good
    total_revenue         -- Actionable: tells you if business is healthy
FROM metrics;

---

Exercises

Exercise 1

What is the primary purpose of a dashboard vs ad-hoc analysis?

Exercise 2

For each potential dashboard metric, identify if it's actionable or vanity:

• Total signups
• Conversion rate
• Page load time
• Revenue
• Active users

Exercise 3

Design the KPIs for a "Cat Cafe Daily Operations" dashboard. What should be on it?

Exercise 4

What questions should you ask before building a dashboard?

Exercise 5

Why might a dashboard with too many metrics be worse than one with few?

---

Key Takeaways

• Dashboard purpose: answer known business questions at a glance
• Build dashboards for decisions, not for data display
• Know your audience and their decisions before designing
• Fewer, well-chosen metrics > many vanity metrics
• Dashboards are for monitoring; analysis is for investigating
• Always connect metrics to decisions they inform