# How explores, dimensions and metrics work

Every chart and dashboard in Kontala Analytics is built from the same three building blocks.

- An **explore** is where a question starts: one subject, such as visits, page views or carts, with its fields ready to use.
- A **dimension** describes something, such as a channel, a country or the time a visit started.
  You group and filter your results by dimensions.
- A **metric** measures something, such as visits, conversions or revenue.
  A metric summarises many rows into one number.

Kontala defines all three for you, in a [semantic layer](/docs/analytics/semantic-layer) it maintains for your organisation, so every chart that uses a metric gets the same number.

## One question, three building blocks

Take the question "How many visits did each channel bring each week, and what share of them converted?"

| Part of the question | Building block | What you pick |
| --- | --- | --- |
| Visits, the subject of the question | Explore | **Visits** |
| How many visits | Metric | **Visits** |
| What share of them converted | Metric | **Conversion rate** |
| Each channel | Dimension | **Channel** |
| Each week | Dimension, at a time interval | **Visit start**, by **Week** |

You open the Visits explore, select the two dimensions and the two metrics, and run the query.
Kontala returns one row for each combination of week and channel:

| Visit start week | Channel | Visits | Conversion rate |
| --- | --- | --- | --- |
| 2026-09-28 | Organic Search | 4,210 | 2.9% |
| 2026-09-28 | Paid Search | 1,875 | 4.1% |
| 2026-10-05 | Organic Search | 4,388 | 3.0% |
| 2026-10-05 | Paid Search | 1,940 | 3.8% |

The figures are illustrative.
To look at recent weeks only, you would add a filter on **Visit start**.

## Dimensions describe, metrics measure

A quick test tells them apart: can you read the value from a single row?

- If you can, it is a dimension: a visit's channel, its landing page, whether it converted.
- If you need many rows to work it out, it is a metric: the number of visits, the conversion rate, the median days to convert.

The same information often appears as both.
**Converted** is a dimension, true or false for each visit, which you can filter or group by.
**Converting visits** is a metric: how many visits converted.

In the Explorer, dimensions are shown in blue and metrics in orange, both in the field list and in the results table.

## Metrics are calculated for each group

A metric has no single fixed value.
It is calculated for every combination of the dimensions in your query.

- **Visits** on its own returns one number.
- Add **Channel** and it returns one number per channel.
- Add **Visit start** by **Week** as well and it returns one number per channel per week.

Filters decide what goes into each calculation.
A filter on a dimension removes rows before your metrics are calculated, for example visits that did not convert.
A filter on a metric removes rows from your results after the metrics are calculated, for example channels with fewer than 100 visits.

## Where the definitions come from

Kontala defines every explore, dimension and metric, each with a description, and keeps them the same for every organisation.
Your Data Capture settings, such as page types and click types, flow into them when someone publishes changes.
[Explores, metrics and the semantic layer](/docs/analytics/semantic-layer) lists every explore and what it covers.

For a one-off question, you can add your own fields to a chart without changing the shared definitions:

- [Bins](/docs/analytics/dimensions#group-numbers-into-bins), which group a number such as **Days to convert** into ranges.
- [Custom metrics](/docs/analytics/metrics#custom-metrics), such as the average of **Pages in visit**.
- [Table calculations](/docs/analytics/metrics#table-calculations), such as each channel's share of all visits.

## Next steps

- [Explores](/docs/analytics/explores): open an explore, then build, filter, sort and save a query.
- [Dimensions](/docs/analytics/dimensions): dimension types, dates and time intervals, filters and bins.
- [Metrics](/docs/analytics/metrics): how metrics are calculated, custom metrics and table calculations.