Most agencies can’t say which of their dashboards anyone still opens. Reports accumulate — one per client, then one per channel, then a one-off somebody asked for in a QBR eighteen months ago — and nothing in the process ever removes one. The work of maintaining them is real; the evidence that they’re being read isn’t.
There is a field in Data Studio that fixes the measurement half of this, and almost nobody fills it in. It takes about four minutes per report. This covers how to set it up, what it genuinely tells you, what it doesn’t, and the three gotchas that quietly switch it off again.
I wrote this technique up back in 2021, when it was a genuinely under-used corner of Data Studio. The feature works the same way today; the one thing that moved is the ID format, which changed for everyone with the GA4 migration. So this is that guide, brought up to date.
Why report usage is worth measuring at all
The case isn’t curiosity. It’s that three expensive decisions all depend on knowing the answer and are currently made on instinct.
What to retire. A report nobody opens isn’t free — it holds a data source, it breaks when a connector changes, and somebody fixes it. Retiring it needs evidence, because retiring it on a hunch means discovering in three months that the finance director loved it.
What to invest in. If the summary page is opened forty times a month and the channel detail pages twice, that tells you where to spend the next hour of build time — and it is usually the opposite of where the last one went.
Whether a redesign worked. Rebuilding a report and then asking the client whether it is better gets you a polite yes. Usage before and after is the version that survives contact with the renewal conversation.
That last one connects to a wider argument: most agencies can’t answer “which of our clients has a dashboard and who maintains it”, and usage data turns that from a filing problem into a list you can act on.
What Data Studio actually measures, and what Google’s own page says
Worth being precise here, because Google’s own documentation says two things that don’t quite line up.
Google’s page on measuring report usage opens by framing three questions the feature answers: how many views your reports are getting, how users are finding them, and where viewers are coming from. That’s the promise, and it’s what you would expect from dropping a Google Analytics tag onto anything.
Then, further down the same page, under Limits of report measurement, it says:
“Google Analytics collects only the
duplicate_reportevent. Data Studio sends this event when a user copies (duplicates) a report.”
Those two statements don’t sit comfortably together. The most likely reading is that the standard page_view hit still fires when someone opens the report — which is what produces the view counts and acquisition data promised at the top — and that duplicate_report is simply the only custom event Data Studio emits on top of that. But Google doesn’t say so, and no other Google-owned page resolves it.
So check it against your own property rather than taking anyone’s word for it, ours included. Paste a measurement ID into one test report, open it in an incognito window, and watch GA4’s Realtime report. Four minutes, and you’ll know exactly what your setup collects.
How to set up report usage tracking in Data Studio
Four steps, and the whole thing lives in one panel.
Step 1: create a GA4 property for your reports
Use a property that exists for this and nothing else. Mixing dashboard usage into a client’s website property poisons that property’s data — your own team opening a report every morning becomes traffic in the client’s numbers, and you’ll spend a year explaining it.
A single property covering all your reports is usually right for an agency, because it lets you compare adoption across clients in one place. One property per client is right when the client needs to see their own usage data, or when data governance says their numbers don’t sit alongside anyone else’s.
Step 2: copy the measurement ID, not the tracking ID
In GA4: Admin → Data streams → your web stream. The measurement ID sits at the top right and begins with G-.
This is the one part of the setup that changed with the GA4 migration, so it is worth stating as a straight before and after.
| What the field took | Where you found it | |
|---|---|---|
| Before | A Universal Analytics tracking ID, UA-000000-2 |
Admin → Property settings |
| Now | A GA4 measurement ID, G-XXXXXXXXXX |
Admin → Data streams → your web stream |
Universal Analytics stopped processing data on 1 July 2023, and Google’s current documentation for this field states that “Google Analytics 4 measurement IDs begin with ‘G-‘”. Everything else about the setup — where the field lives, what it does, how long it takes to populate — is the same as it ever was.
Step 3: open the report settings
In the report, in edit mode: File → Report settings.
The panel that opens holds five things, in this order: the data source, the date range dimension, the report filter, viewer data refresh, and — at the bottom — the Google Analytics measurement ID. It’s the last field in the panel, which is a large part of why so few people have ever noticed it.
Step 4: paste the ID
Paste and close. There is no save button; the setting applies to the report immediately.
Then wait. Google’s documentation says “It may take up to 24 hours before data from your reports begins to show up in Google Analytics”, so an empty report the same afternoon is expected rather than a sign you did it wrong. Realtime is the exception and is the reason to check there first.
Three things that quietly switch it off again
All three are documented, none is obvious, and each one produces a report that looks tracked and isn’t.
A copied report loses its measurement ID
This is the one that matters most to agencies, and it’s stated outright in Google’s documentation on report settings: “If the report is copied, the measurement ID does NOT get copied to the new report.” Google gives the reason on the same page: it is “to protect the privacy of the Google Analytics account owner”, which is fair enough — a template you share publicly should not quietly send its copies’ traffic to you.
Consider what that means for the standard agency workflow. You build a house template, you measure it, it works. Every new client report is a copy of that template — and every one of them arrives with the measurement field blank. The template is the only thing being measured, and the template is the one artefact nobody uses.
The fix is procedural rather than technical: pasting the measurement ID becomes a line on whatever checklist you use when a new client report goes live. Copying and reusing a Data Studio template covers the rest of what does and doesn’t survive a copy, and the measurement ID belongs on that list.
One measurement ID per report
Google’s limits are explicit: “Your reports can have only one Google Analytics 4 measurement ID.”
So you can’t send usage to both your own property and the client’s. You have to choose, and the choice is worth making deliberately rather than by whoever set the report up first. Our default is the agency property, because the question “which of our reports get used” is one you need answered across the whole portfolio, and a client rarely asks how often they opened their own dashboard.
Embedded reports behave differently
If a report is embedded in an intranet page or a client portal, the surrounding page usually has its own analytics, and the two are measuring different things: the parent page records that somebody loaded the page, the report records that somebody loaded the report. They won’t agree, and the gap between them is generally the more interesting number — it’s people who landed on the page and never engaged with the dashboard on it.
Reading the data once it arrives
You now have a GA4 property whose “pages” are your reports. Three views earn their place, and the rest is noise.
Views per report, per month. The blunt one. Sort descending and the bottom of that list is your retirement candidates. Give it three months before acting — quarterly reports open four times a year and will sit at the bottom of every monthly view.
Views around the send date. If the report is scheduled to arrive by email, usage should spike the day it lands. A flat line through the send date means the email is being filed unread, which is a different problem from the report being bad and has a different fix.
Unique users per report. One person opening a report thirty times is a very different situation from ten people opening it three times each, and the monthly total can’t tell them apart. The first is usually the person who built it, checking that it still works.
Name the reports so this is readable. A property full of pages called “Copy of Copy of Untitled Report” is the same governance failure as a Drive folder full of them, and it shows up here first.
What this can’t tell you
Being straight about the limits is what stops the data being over-read in a meeting.
It can’t tell you whether the report was useful. A view is a view; somebody opening a dashboard, failing to find what they needed, and closing it looks identical to somebody getting an answer. If you want to know whether it landed, ask — the reporting call is the place, and usage data is what makes the question specific rather than general.
It can’t tell you which page they looked at with any reliability, because a multi-page Data Studio report is a single-page application and page changes inside it don’t behave like separate page loads.
It can’t tell you who, by name. This is analytics, not an audit log, and it should stay that way — a dashboard that reports on named individuals’ browsing habits is a different and much more sensitive artefact than the one you set out to build.
And it doesn’t survive a redesign cleanly. Rebuilding a report as a new report starts the usage history from zero, which is worth knowing before you use “usage went to nothing” as evidence of anything.
Where this sits, and what it is worth
Report usage tracking is a small, cheap, unglamorous measurement that answers a question most reporting teams have simply agreed not to ask. Four minutes per report, and a standing item on the checklist so copies don’t arrive blank.
And the larger point is the one I closed that original write-up with, which is as true now as it was then:
“When working with clients, always make sure their needs are addressed and their feedback is incorporated in the design of the report to ensure a higher likelihood of adoption across the organization.”
Adoption is the thing being measured. Usage data doesn’t create it — the report having something to say creates it — but it is the only way to find out whether any of it worked.
A note on the Data Studio and Looker Studio names
Google renamed the tool Looker Studio in 2022 and then, in April 2026, renamed it back to Data Studio. Existing reports and links keep working, so nothing needs changing at your end.
One practical note for anyone chasing a source: the documentation moved as well as the name. The current home is docs.cloud.google.com/data-studio/, and help-centre links in the older support.google.com/looker-studio/answer/… format no longer resolve. If a link to Google’s docs gives you a 404, the article almost certainly still exists — search for it from the new docs home rather than assuming it was withdrawn.
Frequently asked questions
How do I track how many people view my Data Studio report?
Create a GA4 property, copy its measurement ID from Admin → Data streams (it begins with G-), then open the report in edit mode and go to File → Report settings. The measurement ID field is the last one in that panel. Allow up to 24 hours for data to appear, though Realtime will show activity immediately.
Why did my report stop being tracked after I copied it?
Because the measurement ID is deliberately not carried over. Google’s documentation states that if a report is copied, the measurement ID does not get copied to the new report. Anyone templating client dashboards is measuring only the template unless pasting the ID is a step on the go-live checklist.
Can I use a Universal Analytics tracking ID?
Not any more. The field originally took a UA tracking ID in the UA-000000-2 format; since the GA4 migration it takes a measurement ID beginning with G-, found under Admin → Data streams. Universal Analytics stopped processing data in July 2023.