Dimensions vs metrics
Who this page is for
| Audience | Why it matters to you |
|---|---|
| Everyone | This is the fundamental reporting model |
| Analysts | Master this to write correct measurement requirements |
Dimensions vs metrics
The foundation of all Adobe Analytics reporting is the dimension-metric pair. Every report is: "[metric] by [dimension]".
Metric — the counter
A metric is a number you want to count. It answers "how many" or "how much".
Examples:
- Visitors (unique people)
- Visits (sessions)
- Page views (times a page loaded)
- Revenue (dollars)
- Form submissions (event fires)
Metrics are always numeric. You sum, average, or deduplicate them.
Dimension — the grouping axis
A dimension is a category you want to group by. It answers "break that down by what".
Examples:
- Page name (break down metric by page)
- Device type (break down metric by desktop/mobile/tablet)
- Traffic source (break down metric by organic/paid/direct)
- Form type (break down metric by contact vs complaint)
Dimensions are text or categorical. You group by them.
The dimension-metric table
Every report is a 2D table: metrics across the top (columns), dimensions down the left (rows).
| Page Name | Visitors | Page Views | Revenue |
|-----------------|----------|------------|---------|
| Home | 1,200 | 3,500 | $0 |
| Product | 800 | 2,100 | $2,500 |
| Checkout | 450 | 900 | $5,200 |
- Each metric is a column (Visitors, Page Views, Revenue)
- Each page is a row (Home, Product, Checkout)
- Each cell is the metric value for that dimension value
Metrics defined by collection
Page views: Fires every time a page loads (you track it with page view rule)
Visitors: Unique ECID values in the report period
Visits: Groups of hits separated by 30-minute inactivity
Revenue: Value of the revenue variable you sent
Form submissions: Count of times event1 fired on your form
Common dimension examples for Example Super
- Page name: / (home), /complaint-form, /search-results
- Form type: contact form, complaint form
- Device type: desktop, mobile
- Referrer: google, direct, linkedin.com
- Search term: what the user typed in site search
- Date: break down by day, week, month
Dimensions from variables
You collect dimension values via:
- Props: Hit-level traffic variables (page section, device type)
- eVars: Session or visitor-scoped conversion variables (campaign source, customer segment)
- Automatic dimensions: Date, device, referrer (Adobe collects these automatically)
Metric arithmetic
You can create calculated metrics from base metrics:
- Bounce rate = bounces / visits
- Revenue per visitor = revenue / visitors
- Conversion rate = orders / visitors
- Average order value = revenue / orders
Dimension-metric mistakes
Mistake 1: Treating a dimension as a metric
- ❌ "How many form types did we have?"
- ✓ "How many submissions by form type?" (form type = dimension, submissions = metric)
Mistake 2: Confusing the rows and columns
- ❌ Putting metrics down the left (wrong orientation)
- ✓ Putting dimensions down left, metrics across top
Mistake 3: Creating unsumming dimensions
- Example: if you report visitor count by timestamp, each timestamp has different visitors
- Sum of all timestamps ≠ total visitors (you overcounted due to visitor overlap)
Reporting example: Example Super
Question: "How many visitors came to the complaint form page, and how many submitted?"
Setup:
- Dimension: Page name (and filter to "/complaint-form")
- Metrics: Visitors, Orders (count of form submissions)
Answer:
| Page Name | Visitors | Orders |
|--------------|----------|--------|
| Complaint | 450 | 125 |
"450 visitors landed on the complaint form page, 125 submitted."