You published a staging review on Tuesday, sent the link to three people, and by Thursday nobody has posted a comment. Silence could mean the build is fine. It could also mean nobody opened the link. The analytics page on your published review is where you tell those two situations apart.
Every review you publish gets an owner-facing analytics page showing visit counts and visitor geography. It is not a dashboard you have to configure. It sits behind the review you already own, and it answers one blunt question: did the people you sent the link to actually open it?
What the numbers actually count
The visit count is the number of times the public review link was opened. A visit is a page load, not a person. If your lead developer opens the link on their laptop in the morning and again on their phone that evening, that is two visits from one reviewer. If they refresh the page four times while reading, that inflates the count too.
So treat the number as a floor, not a headcount. Three visits does not mean three people read it. But zero visits is unambiguous: the link has not been opened at all. That is the reading you care about most before a launch, because it separates a quiet reviewer from a link stuck in an unread inbox.
Visitor geography groups those visits by location. When you send a staging link to a client in Lisbon and a contractor in Manila, and the geography shows visits from both, you have real evidence the link reached both parties. When you expected two regions and see only one, you know who to nudge.
Reading the page before a launch
The analytics page earns its keep in the window between publishing a review and shipping. During a staging site review, you are usually waiting on sign-off from people who are busy and scattered. The visit counts turn a vague waiting game into a specific one.
A practical sequence:
- Publish the review and send the short public link to your reviewers.
- Give it a day. Open the analytics page.
- If visits are still zero, the message did not land. Resend, or send through a channel the person actually checks.
- If visits are climbing but comments are not appearing, people are reading and staying quiet. Ask a direct question on a specific item instead of a general "any feedback?"
- If the geography is missing a region you expected, chase that specific reviewer rather than the whole group.
This matters most when you are reviewing a build for a team spread across time zones and you are never online together. You cannot walk over and ask if someone saw the link. The visit count is the next best thing.
What visit counts cannot tell you
Be honest about the limits. The page does not identify people by name. You see a count and a region, not "Priya opened it twice." If two reviewers sit in the same city, their visits blend together. If someone uses a VPN, the geography may point at the wrong country.
It also cannot tell you how carefully someone read. A visit registers whether the reader studied every pinned comment or glanced at the top and closed the tab. If you need to know that a specific point was understood, the comment thread on that item is your evidence, not the visit count. People with the review link can post a comment on any item, and you can mark each one resolved as it gets addressed. That back-and-forth is where real confirmation lives.
Do not over-read a single number either. Five visits the day you send the link often means one engaged reviewer who kept the tab open and refreshed, not five separate readers. Watch the trend across days and the spread across regions, not one figure in isolation.
Pairing analytics with the rest of the review
The analytics page tells you who arrived. What they do next depends on how the review is built. A review where each screenshot carries a numbered pin on the exact spot you mean gives readers something concrete to respond to, which is why visits turn into comments instead of silence.
If you set up one review with a link for the client and the developer, the visit counts help you see which audience engaged first. And because the same review is available as a public page, a PDF, a Word document, and clean markdown at /r/<slug>/markdown, a reader who prefers a document over the web page still counts as a visit when they open the link before exporting.
Cobalt Capture runs entirely in the browser with no install and no signup, and it is free to use. If you published a review without an account, it is kept for 30 days; sign in with Google or a magic link to claim it and keep the analytics page available past that window.
The one check to run before you ship
Before you tell a stakeholder the staging build is signed off, open the analytics page and confirm the reviewers you counted on actually opened the link. If the visits and geography match the people whose approval you need, your sign-off rests on something real. If they do not, you know exactly who to chase before launch instead of finding out afterward. When you are ready to set the next one up, start a new review and send the link the same day.