# How to analyze your data

## Analyze Data in Daasity: Explores, Dashboards, and Collections

This guide shows how to do ad‑hoc analysis with **Looker Explores**, monitor KPIs with **Dashboards**, and organize your work in **Collections**. It’s written for business users and analysts who want a fast, reliable workflow.

***

### How Daasity Organizes Your Data

Daasity provides a curated semantic layer in Looker so you can analyze accurate, unified metrics without wrangling joins.

* **Unified models (recommended for most reporting):**
  * **UOS – Unified Order Schema:** Direct‑to‑consumer orders, revenue, discounts, refunds.
  * **URS – Unified Retail Schema:** Retail/wholesale performance across retailers and channels.
* **Source‑specific models:** When a vendor/source doesn’t cleanly join to others or you need raw, channel‑level detail.
* **Dimensions vs. Measures:**
  * **Dimensions** are “fields you group by” (e.g., Product Name, Order Date, Traffic Source).
  * **Measures** are calculations (e.g., Net Sales, Orders, AOV).

> **Why isn’t everything in one Explore?**\
> Different data sets have different grains and join rules. Combining them can create duplicates or incorrect totals. Separate Explores keep queries fast and results trustworthy.

***

### When to Use Explores vs. Dashboards

| If you want to…                                        | Use                       |
| ------------------------------------------------------ | ------------------------- |
| Ask a one‑off question or iterate quickly              | **Explore**               |
| Start from a best‑practice template of fields/filters  | **Quick Start (Explore)** |
| Monitor a set of KPIs and trends over time             | **Dashboard**             |
| Make a small change to a dashboard tile and dig deeper | **Explore from here**     |

***

### Pick the Right Explore

Explores are grouped by **subject area**. Start with your business question, then choose the matching subject.

**Common subject areas**

* Orders, Customers, Marketing, Traffic
* Inventory, Returns, Shipping, Subscribers

**Examples**

* DTC revenue trend → **UOS (Orders)**
* Retail sell‑through across accounts → **URS (Retail)**
* Channel‑specific deep dive (e.g., ad platform raw data) → **Source‑specific Explore**

***

### Start in an Explore: Blank or Quick Start

You can begin from a blank slate or use a pre‑curated **Quick Start** with common fields, filters, and visual defaults.

**Blank Explore (full control)**

1. Open **Explore** and select the model (e.g., **UOS** or **URS**).
2. Add **Dimensions** (e.g., Order Date) and **Measures** (e.g., Net Sales).
3. Apply filters, click **Run**.
4. Choose a visualization and tweak formatting.

**Quick Start (faster)**

1. Open **Explore** and pick the model.
2. Select a **Quick Start** (e.g., “Orders by Day (Net)”).
3. Review the pre‑loaded fields/filters; adjust if needed.
4. Click **Run** and customize the visualization.

> **Best practice:** Start with a Quick Start for common questions; switch to a blank Explore for bespoke analyses.

***

### Dashboards = Collections of Tiles (Each Tile Is a Saved Explore View)

A **dashboard** is a page of **tiles**. Each tile is backed by an individual **saved Explore view** (a specific query + visualization).

* Use dashboards for **monitoring** and sharing KPIs.
* Each tile can be opened for deeper analysis with **Explore from here** (see below).
* You can **clone** dashboards or add your own tiles from saved Explore views.

***

### “Explore from Here” (Adjust Any Dashboard Tile Safely)

When a dashboard tile raises a new question:

1. On the tile, open the **⋯** menu and choose **Explore from here**.
2. You’ll land on the underlying Explore with the tile’s fields/filters pre‑applied.
3. Adjust fields, filters, pivots, or visualization.
4. **Save** your adjusted analysis (see **Collections** below). You won’t change the original dashboard tile unless you explicitly update it.

> **Tip:** Use “Explore from here” to keep the dashboard stable while you iterate in your own workspace.

***

### Work Faster: Pivots, Filters, and Visualizations

#### Pivoting

* Hover over a **Dimension** and click **Pivot** to turn its values into columns.
* Keep pivoted columns **under \~50** for snappy browser performance. (Up to **200** pivoted values supported; visibility often limited for performance.)
* Sort columns by clicking headers; **Shift + click** for multi‑column sort.
* To unpivot, use the gear menu (**Unpivot**) or click the Pivot icon again.
* Include at least **one unpivoted Dimension** and **one Measure**.

#### Filtering

* **Basic filters:** Use the filter icon next to any field in the picker.
* **Advanced:** **matches (advanced)** supports pattern matching for complex expressions.
* **Custom Filter:** Build compound logic using Looker expressions.
* **Behavior:**
  * **Dimension filters** limit the raw rows considered.
  * **Measure filters** apply **after** aggregation (e.g., only show products where Net Sales > X).
* **Limits:** Set row/column limits to keep queries responsive (typical defaults: up to **5,000 rows** and **200 columns**).

#### Visualization

* Quick‑switch among **Table, Column, Bar, Line, Pie**, etc.
* Click **Edit** to configure axes, series, labels, number formatting, and totals.
* For tables, add row/column **Totals** and drag headers to **reorder** columns.

> **Performance tip:** Prefer fewer, more targeted fields; use filters and modest pivots to keep results fast and readable.

***

### Quick Decision Guide: Choosing the Correct Explore

1. **Define the question:** Revenue trend? Cohort behavior? Traffic source? Inventory position?
2. **Match the subject area:** Orders, Customers, Marketing, Traffic, Inventory, Returns, Shipping, Subscribers.
3. **Pick unified vs. source‑specific:**
   * Use **UOS/URS** for standardized, cross‑source reporting.
   * Use **source‑specific** when you need vendor‑level details that don’t belong in a unified model.

***

### FAQs

**Can I customize a default dashboard tile?**\
Yes. Use **Explore from here**, make changes, and **Save** to your **Private** or **Shared** Collection. You can also clone the dashboard and replace tiles.

**Where should I save finished analyses?**\
Place team‑relevant, stable content in a **Shared** subcollection. Keep drafts in **Private**.

**What if I don’t see an Explore I need?**\
Some models are permissioned or require enablement. Contact your admin or Daasity Support.

***

### What to Read Next

* **Accessing Explores** – Find and open Explores in your environment
* **Scheduling & Alerts** – Send results to Slack/email on a schedule
* **Permissions & Sharing** – Control who can view or edit your assets

> **Remember:** Use **Explores** to answer questions quickly, **Dashboards** to monitor what matters, and **Collections** to keep everything organized and shareable.