# Explorations

AdPix's free-form table builder: put dimensions in the rows, metrics in the columns, pivot one dimension across the top, then save the result or export it. This is a premium feature and is not in the Free plan by default.

## What this page is

Each prebuilt AdPix report answers one fixed question. Explorations are for the question that has no prebuilt report: "of the mobile users who came from the Nowruz campaign, which cities produced the most key events?" You assemble the dimensions and metrics yourself and build the table.

The page is at `/explore` and appears in the left navigation under the **Explore** group as **Reports & Explore**.

> **This is a premium feature**
>
> Reports & Explore is on the **Premium** plan by default and is not included in the Free plan. The plan belongs to the **property**, not to you — so the feature can be open on one of your properties and locked on another. When it is locked, the page still opens but renders "Reports & Explore is a premium feature" with an **Upgrade plan** button, and the navigation item carries a premium marker. Which features sit in which plan is set by AdPix and can change.

## The two modes

A toggle at the top of the page switches between them:

- **Reports** — eight prebuilt cards for common questions: Cities × Devices, Channel performance, Landing pages, Campaigns, Daily trend, Source / Medium, Countries and Tech breakdown. Selecting a card opens that combination in the builder so you can refine it. Below the cards, this property's **Saved explorations** are listed too.
- **Explore** — the builder itself, in three columns: variables, settings and the canvas.

## Build a free-form table

1. Set the toggle at the top of the page to **Explore**.
2. In the **Dimensions** column, select the dimensions you want. Three at a time is the maximum; picking a fourth drops the oldest one.
3. In **Metrics**, select up to four. Order matters — the table is always sorted by the first one.
4. To turn a dimension into columns instead of rows, put it in **Columns (pivot)**.
5. Pick **Table**, **Bar**, **Line** or **Donut** under **Visualization**, and a row count under **Show rows**.
6. There is no Run button. Every change rebuilds the table after a short pause. **Save** names it; **Export CSV** gives you exactly what is on screen.

> Building a free-form exploration: dimensions, metrics and segments. — [analytics.adpix.io/en/explore](https://analytics.adpix.io/en/explore)

## The dimensions you can use

| Dimension | What it groups by |
| --- | --- |
| Date | the day of the event, bucketed in the property's timezone |
| Event name | the event's name, such as `page_view` or `purchase` |
| Page path | the path of the page the event happened on |
| First-touch channel | the channel that first brought the user |
| Source, Medium, Campaign | the user's first-touch source, medium and campaign |
| Last-touch channel | the last channel seen before the event |
| Country, City | where the event came from |
| Browser, OS, Device category | the device's technology |

Any event parameter you have registered as a custom dimension on the property joins the same list. Registering them is covered in [custom definitions](analytics/collect/custom-definitions).

> **First-touch dimensions are user-scoped**
>
> Source, Medium and Campaign in this list describe the user's **first** touch, not the touch that brought the session you are looking at. If your question is "what brought traffic this month", [traffic acquisition](analytics/reports/traffic-acquisition) answers it more accurately, because that report is session-scoped.

## The metrics, defined exactly

| Metric | What is counted |
| --- | --- |
| Users | distinct visitor IDs. One person on two browsers is two users. |
| Identified users | only identities that came from a sign-in, an email, a phone number or an external ID. Every visitor gets an internal ID on arrival; those visitor-seeded ones are excluded. |
| Sessions | distinct sessions. Server-side events have no session and are not counted here. |
| Events | distinct events, counted by event ID. |
| Views | page-view events only. |
| Key events | events whose name is marked as a key event on the property. |
| Key event value | the sum of the `value` parameter on those key events. |

Every numeric custom parameter also produces two metrics: its sum and its average.

The key-event list comes from the property's settings — if you have not marked any event as a key event, those two columns are always zero. See [key events](analytics/collect/key-events).

## Narrowing with a segment

Below the **Dimensions** column there is a **Segment** picker. The default is **All users**; the other entries are the segments saved on this property. A segment applies to the whole table, not to a single column.

The segment's scope decides what survives the filter: a user-scoped segment keeps **every** event belonging to any user who had at least one matching event, a session-scoped one keeps whole matching sessions, and an event-scoped one keeps only the matching events themselves.

Segments can currently only be created on the Segments page, which is still parked under **Coming soon** — its state is described in [segments](analytics/explore/segments).

## The pivot, and the caps that really apply

Read this part carefully; most "why is my number different" questions come from here.

- **Three dimensions, four metrics, maximum.** Selecting more drops the oldest selection.
- **Sorting is always descending on the first metric**, and the table is cut to the row count you asked for. You are always looking at a top list, not at every possible combination.
- **In pivot mode only the first metric fills the cells.** The other metrics are not displayed in that mode.
- **Pivot columns are capped at the top six values.** The page says so under the picker: it pivots the metric across the top 6 values of that dimension. Anything outside those six gets no column at all.
- **The Total column still counts everything**, including the values that got no column. So the cells of a row usually add up to less than that row's total — that is not an error.
- **Pivot mode reads more raw data** — up to twelve times the row count, capped at five thousand raw rows — and then aggregates it in your browser. On very high-traffic properties, the long tail of combinations may not fit inside that cap.
- **Charts are coarser than the table.** Bar and donut take the first fifteen rows (the donut draws eight slices), and the line chart draws the first three metrics over the first sixty rows. The exact numbers are always in the table.

> **Start from a prebuilt card**
>
> The fastest way to a correct exploration is to open the closest card in **Reports** and change one or two things in the builder. The cards already combine the dimensions and metrics people actually use together.

## Saving, reloading and exporting

**Save** asks for a name and stores the exploration on the current property. Saved explorations belong to the property, not to you personally, so your teammates see them under **Saved explorations** in the **Reports** mode and open them in the builder with one selection.

What is stored: dimensions, the pivot dimension, metrics, the segment and the visualization. What is **not** stored: the date range. The range always comes from the picker at the top of the page, and changing it reloads the page and re-runs the same exploration over the new range. The picker is covered in [date ranges and comparison](analytics/reports/date-ranges-and-comparison).

**Export CSV** gives you exactly the table on screen — the same columns, the same order, the same number of rows. If you want a fuller export, raise the row count first.

Whatever you last built is remembered in your own browser, so returning to the page reopens the builder where you left it.

## Small things people trip on

- If your combination has no data in the range, you get "No data for this combination in range." Usually that means the dimensions are too fine, or the segment excluded everything.
- Test-mode events are never counted in an exploration.
- The Date dimension and both range boundaries are evaluated in the property's timezone.
- Server-side events carry no session ID and no page path. They are present in Events, Key events and Key event value, but not in Sessions and not in page-based dimensions.

## Frequently asked questions

### Why do I get a premium message instead of the builder?

Because Reports & Explore is a premium feature and the plan belongs to the property, not to your account. If the property selected at the top of the page is on the Free plan, you get "Reports & Explore is a premium feature" and an Upgrade plan button instead of the builder. Check the property switcher first — you may be on a different property than you think.

### Why does my table only have a handful of rows?

Because the table is always a top list, never the complete list. You choose the row count in Show rows, between ten and two hundred and fifty, and rows are ranked by whichever metric is first in Values. To see more of the tail, raise the row count or narrow the data with a segment.

### What is the difference between Users and Identified users?

Users counts distinct visitor IDs, so one person on two browsers is two users. Identified users counts only people whose identity came from a real sign-in, email, phone number or CRM ID — every visitor is given an internal ID on arrival, and those visitor-seeded ones do not count.

### Does a saved exploration remember the date range?

No. A save stores the dimensions, the pivot dimension, the metrics, the segment and the visualization. The date range always comes from the picker at the top of the page, so a saved exploration re-runs against whatever range is active.

## Related

- [Path exploration](https://docs.adpix.io/en/analytics/explore/path-exploration/)
- [Segments](https://docs.adpix.io/en/analytics/explore/segments/)
- [Custom definitions](https://docs.adpix.io/en/analytics/collect/custom-definitions/)
- [Date ranges and comparison](https://docs.adpix.io/en/analytics/reports/date-ranges-and-comparison/)

---

[Docs](https://docs.adpix.io/en/analytics/explore/explorations/) · AdPix
