> ## Documentation Index > Fetch the complete documentation index at: https://openrush.com/docs/llms.txt > Use this file to discover all available pages before exploring further. # Query your website analytics > All the ways OpenRush can read your connected Google Analytics (GA4) data. OpenRush can read your own Google Analytics 4 data once you connect a property. This is measured data from Google, not an estimate. It reports the real sessions, users, engagement, and traffic sources your site earned, read straight from your GA4 property through Google's official APIs. OpenRush reads aggregate data only. It never reads user level data or personal information, and it never writes to your property. One tool, many angles on your website traffic. ## The tool Everything below runs through a single tool: `get_website_analytics`. You do not need to memorize it. Just ask your agent a normal question about your website traffic, and it picks the right settings for you. This tool needs a connected Google Analytics property. Connect one in [Connections](https://www.openrush.com/dashboard/connections). If you have several connected sites, name the one you mean. ## What it measures Every view returns the same core metrics for your site or for each row you break down by: | Metric | Meaning | | ----------------------- | ----------------------------------------------- | | Sessions | Visits to your site | | Users | People who visited (total and new) | | Engaged sessions | Visits that stuck around or converted attention | | Engagement rate | Share of visits that were engaged | | Average engagement time | How long an average visit stayed active | | Page views | Pages and screens viewed | | Bounce rate | Share of visits that were not engaged | These are the questions GA4 measures well for every website: how much traffic you get, where it comes from, who it reaches, and how engaged those visits are. ## The dials Every question is a mix of a few choices. Change them and you get a different view of the same traffic. ### 1. Report What to break the traffic down by. Leave it off for the rich overview, which is site totals plus your top channels, top landing pages, and a daily trend in one call. | Value | Answers | | ------------------ | --------------------------------------------------------- | | `overview` | Site totals, top channels, top pages, and trend (default) | | `channel` | Which marketing channels bring you traffic | | `source_medium` | The exact source and medium behind the traffic | | `source` | Which sources send visits | | `medium` | Which mediums send visits | | `campaign` | Which UTM campaigns drove traffic | | `landing_page` | Which pages visits arrive on | | `country` | Which countries your visitors are in | | `city` | Which cities your visitors are in | | `device` | Desktop vs mobile vs tablet | | `browser` | Which browsers visitors use | | `operating_system` | Which operating systems visitors use | | `date` | How traffic trends over time | ### 2. Time range How far back to look. | Value | Meaning | | ---------------- | ------------------------ | | `last_7_days` | The last week | | `last_28_days` | The last month (default) | | `last_3_months` | The last quarter | | `last_6_months` | Half a year | | `last_12_months` | A full year | ### 3. Compare to Whether to show the change against a past window. | Value | Meaning | | ----------------- | --------------------------------------- | | `previous_period` | Versus the window just before (default) | | `previous_year` | Versus the same dates last year | | `none` | No comparison | ### 4. Filter Narrow to a slice before breaking it down. You can filter on channel, source, medium, source and medium, campaign, UTM content, UTM term, landing page, country, city, device, browser, or operating system, and match by exact, contains, begins with, ends with, or regular expression. You can also set how many rows come back. ## How many ways is that? You can leave the overview as is or break down by any of twelve dimensions, which is thirteen shapes. Multiply that by five time ranges and three comparison modes, and you get close to two hundred distinct views, before you add filters across thirteen more dimensions. In short: if a question about your Google Analytics traffic can be answered from sessions, sources, geography, device, or engagement, this tool can shape the query for it. ## Things you can just ask * How is my website traffic doing this month compared to last? * Which channels send me the most engaged visits? * What are my top landing pages, and which ones are losing traffic? * Is my traffic more mobile or desktop? * Which countries are growing for me year over year? * Which campaigns drove the most sessions last quarter? * How has my daily traffic trended over the last three months? ## Good to know * Attribution is session scoped. Each visit's source is the channel, source, medium, and campaign of that visit itself, and every response says so. * Every window is computed in your property's own reporting timezone, so your days line up with what you see in Google Analytics. * The most recent day or two is marked preliminary. GA4 takes up to a day or two to fully process recent data. * Sessions are visits, not users, and they are not the same as Search Console clicks. Use [search performance](/docs/guides/search-console) for measured Google Search clicks. * When a report is based on a sample of your sessions, or when Google hides rows for small groups of people to protect privacy, OpenRush says so in the response and lowers its confidence rather than showing a number that looks more precise than it is. * High cardinality reports, such as landing pages, roll their long tail into a single `(other)` bucket, so the visible rows can add up to less than the total. OpenRush discloses this. * This is the universal traffic, attribution, and engagement tool. It deliberately reports no conversions, key events, revenue, or custom events, because those are meaningful only for some sites and are easy to instrument badly.