Why your ad dashboard can't tell you which pages earn
AdSense is good at telling you how much you made yesterday. It's much worse at telling you which articles made it. Its reports break earnings down by site, ad unit, country, platform and channel, and you can set up URL channels to track specific pages, but you have to create each channel yourself, and it doesn't scale to a blog with a few hundred posts.
Google Analytics 4 fills that gap. Once you link your AdSense or Google Ad Manager account to a GA4 property, ad revenue arrives as events tied to the page where each ad was shown. That means you can put revenue next to page views, sessions and landing pages, and work out what each piece of content is actually worth. This post walks through the link, the reports and metrics you get, how to build a page RPM and session RPM view, and the gotchas that make GA4 numbers disagree with your AdSense dashboard.
Step 1: Link your ad account to GA4
There are two integrations publishers use. Which one applies depends on how you serve ads.
AdSense
The link is created from the AdSense side. According to Google's AdSense Help article on linking a GA4 property, you sign in to AdSense, go to Account → Access and authorization → Google Analytics integration, click New link, pick the property and create the link. You need Admin access in AdSense and the Editor role on the GA4 property. The property must collect data with the Google tag (gtag.js); Google's Analytics Help notes that older "connected site tags" aren't processed for AdSense measurement, so if you're still relying on those, switch to the Google tag first.
Once it's done, the link shows up in GA4 under Admin → Product links → Google AdSense links. Google says data can take up to 24 hours to start appearing, and AdSense data from before the link existed is never back-filled. That last point matters: if you link today, your page-level history starts today.
Google Ad Manager
If you serve through Ad Manager (directly, or through a partner who gives you access to your own network), the link lives under Admin → Product links → Ad Manager links in GA4. Google's Analytics Help lists the requirements: you need to be an Ad Manager administrator and an Analytics Editor, the site must use the Google tag, and most of your ad tags need to be Google Publisher Tags (GPT). Ad Manager tags served inside iframes, including most passbacks, don't share data with GA4.
One detail worth knowing: GA4 calculates Ad Manager revenue from the line item's rate at the moment the impression is logged. If the rate changes later, or billing is reconciled afterwards, the GA4 figure doesn't change. Treat it as a very good estimate, not an invoice.
If you use a managed network that runs its own wrapper, like the ones covered in our network comparison, check with them. Some pass revenue into GA4 through their own integration or through the ad_impression event; others only report in their own dashboard. Google doesn't control that, so the network's documentation is the source to trust.
Step 2: Find the publisher ads data
With the link active, GA4 collects ad_impression and ad_click events automatically for each ad on a page (it also uses an internal ad_query event to join the two systems). Google's help pages currently place the Publisher ads report under Advertising → Publishing → Publisher ads; some older AdSense help text still refers to a Monetization section, so if you can't find it in one place, look in the other. If the report is missing entirely, an Editor or Administrator can add it back to the navigation through the report library.
The report uses four metrics:
- Publisher ad impressions: ad impressions recorded through the integration.
- Publisher ad clicks: clicks on those ads.
- Ad unit exposure: how long ad units were on screen.
- Total ad revenue: revenue from your linked ad integrations.
And these dimensions: ad format, ad source, ad unit, and page path and screen class. That last one is the useful bit. It's what lets you see revenue per URL.
The built-in report is fine for a glance. For decisions, use an exploration.
From account link to content decisions
- Link AdSense or Ad ManagerAdSense side, or GA4 Admin → Product links
- Wait and collectUp to 24 hours to start; no history before the link
- Build an explorationPage path rows; views, sessions, total ad revenue
- Calculate page RPM and session RPMRevenue ÷ views or sessions × 1,000
- Sort into update, expand, pruneUse 60–90 days of data, not one week
Step 3: Build a page RPM and session RPM view
GA4 doesn't ship a "page RPM" metric, so you make one. There are two ways.
Option A: a free-form exploration
- Go to Explore and start a blank free-form exploration.
- Import the dimensions Page path and screen class and Landing page + query string.
- Import the metrics Views, Sessions and Total ad revenue.
- Put page path in Rows and the three metrics in Values. Sort by total ad revenue.
- Set the date range to at least the last 60 to 90 days, so one viral day doesn't skew everything.
- Export to Google Sheets and add the RPM columns there.
Option B: a calculated metric
GA4 lets you create calculated metrics under Admin → Custom definitions, using predefined metrics in a formula. A standard property gets five of them (360 properties get 50), so they're precious, but "Total ad revenue ÷ Views × 1000" is a good use of one. Once it exists, you can drop it straight into an exploration or a custom report without exporting.
The two ratios, and what they mean
Page RPM is total ad revenue divided by views, times 1,000. It tells you how well a single URL monetizes each time someone loads it.
Session RPM is total ad revenue divided by sessions, times 1,000. Use it with the landing page dimension and it answers a different question: when a visitor arrives on this page, how much revenue does the whole visit generate, including the other pages they go on to read? A post with mediocre page RPM can have strong session RPM because it sends readers deeper into the site.
The table below shows how the same set of numbers reads differently through each lens. The figures are made up for illustration; they aren't benchmarks.
Metric, formula, and what it tells you (hypothetical numbers)
| Metric | Formula | Example | What it tells you |
|---|---|---|---|
| Page RPM | Total ad revenue ÷ views × 1,000 | $30 ÷ 10,000 views × 1,000 = $3.00 | How well this URL earns per load |
| Session RPM (by landing page) | Total ad revenue ÷ sessions × 1,000 | $45 ÷ 6,000 sessions × 1,000 = $7.50 | What a visit starting here is worth |
| Views per session | Views ÷ sessions | 10,000 ÷ 6,000 = 1.67 | Whether readers keep going |
| Revenue share | Page revenue ÷ site revenue | $30 ÷ $600 = 5% | How much depends on this page |
If you want to sanity-check what an RPM change would mean in monthly terms, plug your traffic into the revenue calculator. For the levers that move RPM itself, see how to raise ad RPM.
Step 4: Turn the numbers into content decisions
Once you have page RPM and session RPM side by side, most pages fall into one of a few groups. A rough way to sort them:
- High traffic, low page RPM: the audience is there but the page doesn't monetize. Often it's short, has one ad slot, or the ads sit below where people stop reading. Look at placement and whether the page actually answers the query in enough depth to hold attention. Update these first; small RPM gains on big pages add up fastest.
- Low traffic, high page and session RPM: the topic pays well and readers engage. These are candidates to expand: write related posts, link to them internally, improve the page so it can rank for more queries.
- Low page RPM, high session RPM: a gateway page. It earns little itself but sends people onward. Don't prune it on page RPM alone.
- Low traffic, low RPM, low session RPM, for months: the prune-or-merge pile. Before deleting, check Search Console for impressions; a page with search demand may just need a rewrite. If it's genuinely thin and overlaps another post, merge it and redirect. Our content quality guide covers what "thin" looks like to a reviewer.
One honest caveat: revenue isn't the only reason a page exists. Your about page, contact page and privacy policy will all show near-zero RPM. They're still required for trust and for most network approvals, so leave them out of the pruning exercise.
Pitfalls that make GA4 disagree with AdSense
Your GA4 total ad revenue won't match your AdSense or Ad Manager dashboard exactly. That's expected. Here's why, and what to do about each.
Different time zones
Google's Analytics Help says that if your AdSense account and GA4 property use different time zones, the same calendar date aggregates different hours, so daily totals won't line up. You can change the GA4 reporting time zone in property settings. Matching the two makes day-by-day comparisons far less confusing, but changing it only affects data going forward.
Ad blockers and tags that don't fire
Google lists security and ad-blocking software as a cause of discrepancies: a visitor can be counted in GA4 while the ad never loads, or the analytics tag can be blocked while ads still serve. If GA4's tag doesn't execute on a page, the AdSense data for that page has nothing to attach to and doesn't show up in GA4. The same goes for pages where the Google tag's config command doesn't run before other events. Make sure the tag loads on every template.
Revenue is attributed to the page where the ad showed
GA4 credits revenue to the page path where the impression happened, not to the article that brought the reader in. A strong entry article that sends people to a high-RPM page gets no page-level credit for that. That's exactly why session RPM by landing page is worth building alongside page RPM. Also watch query strings: if your URLs carry tracking parameters, use page path without the query string, or one article can split into dozens of rows.
Estimates, not finalized earnings
AdSense reports estimated earnings that can be adjusted when the month is finalized, for example after invalid traffic is removed. GA4 records what it received at the time. Use GA4 for comparing pages against each other, and your AdSense or Ad Manager dashboard for what you'll actually be paid. Google makes a similar point about Ad Manager: billing reconciliation doesn't flow back into GA4.
Sampling and thresholds
Explorations can be sampled when a query covers more events than the property's quota, which Google sets at 10 million events per query for standard properties. A small or mid-sized blog will rarely hit that, but if you see the sampling indicator, shorten the date range. Data thresholds are a separate thing: GA4 may withhold rows to protect user privacy, most often when Google signals or demographics are involved and counts are low. Google says thresholds are system-defined and can't be switched off. Keeping demographic dimensions out of your RPM exploration and using a wider date range usually avoids them.
Consent mode reduces what you can see
If you use consent mode and a visitor denies analytics storage, the Google tag doesn't set analytics cookies and sends cookieless pings instead. GA4 can model the missing users and sessions through behavioral modeling, but only when the property meets Google's eligibility requirements, and Google notes that modeling isn't applied to event counts like page_view. In regions where many visitors decline, your views and sessions may be undercounted relative to ad revenue, which inflates RPM. Compare RPM within the same region rather than mixing EU and non-EU traffic in one number.
A quick checklist before you trust the numbers
- The Google tag is on every page template and fires its config command first.
- The link appears under Admin → Product links in GA4.
- The AdSense or Ad Manager time zone matches the GA4 property time zone.
- You're using at least 60 to 90 days of data since the link was created.
- Your exploration uses page path without query strings.
- You've read page RPM and session RPM together before deciding anything.
If you're still waiting on approval and don't have ad revenue to track yet, the first job is getting the site ready. Run a free scan on the AdGateScore homepage to see how your site lines up against each network's requirements.
We build AdGateScore, the ad network readiness scanner. Our articles draw on the requirements each network publishes and on what our scans find.