Know who came to your site, where from and what they did, without opening GA4.

Where did they come from, and what did they read?

Traffic by channel splits sessions the way GA4 groups them: organic search, paid search, direct, referral and organic social, up to eight channels with a count each. Under it, source / medium lists up to twelve rows, so google / organic and google / cpc are separate lines and a referral from one site has its own row. A devices table splits the same sessions across mobile, desktop and tablet. With a comparison on, every row carries the prior figure and the change, coloured by direction.

Top pages ranks the pages people read by views. An engagement table under it carries bounce rate, average session length, engaged sessions and new users, this period beside the comparison. Read the two together. A service page that draws its views from organic search and holds a reader for three minutes is doing its job. A blog post that draws views from social and loses readers in twenty seconds is a post, not a lead source. This is the card that says which is which.

The demo shows the shape with illustrative numbers, and says so. Its Analytics view reads 1,740 sessions in the latest month, up 7 percent on the month before. Organic search carries 38 percent of traffic, AI referrals 16 percent and direct 15 percent. The top page holds readers for 3 minutes 12 seconds, and 84 demo requests were counted in the last 30 days. Your instance reads your property, and only the demo invents a figure.

Why is GA4 hours behind when my CRM is live?

Because Google publishes GA4 intraday data on its own schedule, hours after the fact, and no dashboard can read what Google has not released. The page states the ceiling per source instead of labelling everything live. Your CRM, Stripe, your booking platform, website leads and email stats are live, and a refresh moves them. Google Ads lags hours. GA4 intraday lags hours. Search Console releases each day two to three days later.

Here is what happens when you open the page. Last night's snapshot renders first, so there is something to read before any network call returns. The live payload comes from a cache at once. If that snapshot is older than five minutes, a pull runs behind the request and the numbers update when it lands. The header stamp says how old the data is, and its hover text says which case you are in: pulled live when you opened the page, or the last saved snapshot because the live refresh was unreachable. Press Refresh and a pull of two seconds to a little over three minutes starts; the page reloads when the stamp moves, and every path through it ends in a reloaded page.

Two things run on a schedule whether anyone is at a desk. At 12:01am Central, one job rebuilds the fallback snapshot for every client from one registry, and a pull that fails or a file that fails its syntax check leaves the previous page up, so the page you open is always a whole one. At 7:20am Eastern, the lead watch runs where it is set up. Two things stay with a person. Which events count as key events is set in your own GA4 admin. A change you request from the screen, to content, keywords, categories or targeting, goes to the team as a logged ticket, and the engine's writes to your ad accounts and your site go through each platform's API and land in its own change history.

Know the edge of the view as well. It reads your GA4 property; the rest of your dashboard reads the sources connected to your instance, with more connected during setup. Your customers' card data is never read. A phone call counts when your CRM records it. Google AI Overviews are measured on the search visibility view.

How do I connect it, and what should I check in the demo?

Add our Google account as a manager on the property, from access management in your own GA4 admin. That is the whole connection: our team does the rest during setup, and we never ask for your account password. You grant the access during setup and revoke it in the same screen at any time.

On our side the property is saved once in your account's connection map, and every request the dashboard makes uses that saved property, so the property id stays on the server, and the dates the browser sends are checked before they reach Google. Revoke the access and the next window that has not been pulled answers with an unavailable card instead of a number; the rest of the dashboard is unaffected, and the nightly bake leaves the last good page up.

Check it in the demo before you buy. Open demo.trulata.com, go to Analytics and click the sessions card: the drill-down shows the latest month, the change on the month before, and names GA4 as the source. Click a traffic source for its share, a page for its views and average time, and a conversion tile for its 30-day count. The demo is a fictional company on the same engine code every client runs, with your name and email in front of it. The date picker is switched on per instance during setup, so the demo shows the twelve-month view.