eoma

Docs › Screens

Web analytics

Your real website traffic from Google Analytics — who arrived, where they came from, and how many of them were sent by an AI assistant.

In 10 seconds

Before you start

Steps

1 — Open it from the sidebar

┌─ sidebar ─────────────────────────┐
│  ▶ SEO                            │
│  ▼ Web Analytics                  │
│    ● Google Analytics    ← here   │
│      Search Console               │
│  ▶ Social Listening               │
└───────────────────────────────────┘

Web Analytics is its own section holding two screens. This page is the first of them; the second is Search Console.

It also sits on the dashboard orbit map as the Analytics planet.

2 — Choose your reporting window

┌───────────────────────────────────────────────┐
│ Web Analytics            [24h][48h]│[7d][30d][90d] │
│ example.com · 7d overview             ↑ pick one  │
└───────────────────────────────────────────────┘

The line divides two kinds of window. 24h and 48h report hour by hour; the rest report day by day. That changes the grain of every number below, which is why the divider is there.

3 — If you have more than one property, pick it

When your Google account has several properties linked to this brand, the site name under the title becomes a dropdown. One property, and it is plain text.

What you'll see

Blocks appear only when they have data, so a small site shows a shorter page. Nothing is hidden from you deliberately.

The header — "Web Analytics"

Sticky at the top. Under it: the property name, the window you picked, and — on the 24h and 48h windows only — which timezone the hours are counted in. That last part matters: "nothing since 14:00" is a different fact in Istanbul than in Berlin.

The four tiles — Active Users, Sessions, Bounce Rate, Avg Duration

The headline row, each with a small trend line and a comparison to the window before. Revenue tiles join them if your site reports sales to Analytics, then a row carrying a tag for how Analytics reads your site — E-commerce, Lead Gen or Content — and any sharp movement worth noticing.

"Sessions & Users" and the live panel

The main chart on the left. On the right, Live: who is on your site in the last 30 minutes, with the pages, sources and countries they came from.

"Where Visitors Came From"

The widest table on the page, and usually the reason people open it. One row per source, with a bar strip showing when they arrived, a session count, and how long ago the last one was. A row can carry an AI tag, and a line above the table totals those.

"Traffic Sources" and "Top Pages"

Side by side. The channel mix on the left is clickable — clicking a channel filters the table above it to the sources behind it. Under them, "Devices" and "Countries", then "New vs Returning Users" and "Landing Pages".

Conditional blocks

Campaigns, Site Search Terms, Google Ads, Ecommerce and Conversion Funnel load a moment after the rest, and only when your Analytics property reports them.

Reading the numbers

Active Users vs Sessions — for example 1,240 users and 1,830 sessions. 1,240 different people visited, and between them they came 1,830 times. The gap is repeat visits. Users going up while sessions stay flat means you reached more people who each looked once; the reverse means the same audience came back more often.

Bounce Rate — for example 41.2%, and Avg Duration — 1m 48s. Four in ten visits ended without the person doing anything Analytics counts as engagement, and the average visit lasted under two minutes counting everyone who left instantly. On the bounce tile, down is good and the colour is inverted to say so. Read the two together rather than alone: both moving is a real change in visitor quality, one moving alone is usually just a change in traffic mix.

"142 sessions came from AI assistants". 142 visits arrived from a link inside an AI answer — someone asked ChatGPT or Perplexity something, got your site as a source, and clicked. This is the closest thing you have to a direct receipt for the AI-visibility work in the rest of EOMA. Treat it as a floor rather than a full count: not every assistant identifies itself when it sends a visitor, so the true figure is at least this.

"Direct / no referrer". No source was recorded. Some of it is people typing your address, some is a browser stripping the referrer, and some is an app that does not pass one. It is the one row you cannot draw conclusions from.

When it goes wrong

You see Why What to do
"Connect Google Analytics" with a three-step list Analytics has never been linked for this brand. Follow the three steps, or read Google Analytics first.
"Finish Connecting Google Analytics" The Google sign-in started but the analytics permission was never approved, so no lasting access was granted. Run the connection again and approve every permission it asks for.
"Pick a Property" Your Google account is linked but no property is chosen for this brand. Open Settings → Integrations and pick the property that tracks this website.
"Reconnect Required" Google revoked or expired the authorization — most often a password change or a permission removed on Google's side. Reconnect from Settings → Integrations. Nothing is lost; collection resumes.
"No Access to This Brand" Your EOMA account cannot view this brand's analytics. Reconnecting will not fix it. Ask whoever owns the brand to add you.
"Couldn't Load Analytics" with a Retry button A transient failure — network or a Google hiccup. Press Retry. It usually works the second time.
You came looking for search keywords and there are none here Search Console moved to its own screen; this page is Analytics only. Open Search Console in the same sidebar section.
The whole page is thinner than yesterday Blocks render only when they have data in the window you picked. A short window on a quiet site drops several. Switch to 30d or 90d before concluding anything is missing.
Hour-by-hour numbers look shifted Analytics counts hours in the property's timezone, which may not be yours. The header names it on those windows. Read the timezone note in the header before comparing to your own clock.
Reading this as a machine? The same page in Markdown: /docs/pages/web-analytics.md

On this page