← All docs

Tracked clicks and page journeys

Tracked clicks is the drill-down behind the Tracked clicks number in the Link analytics card on Analytics. Click that number and you land on a list of every tracked visit in the date range, one row per visit, newest click first. Each row opens to show the pages that visitor reached, in order, with times.

This page is on the Growth plan and above. On the free plan it loads but shows "Click tracking is on the Growth plan" with a link to Plans instead of the list. Collection still runs on every plan, so if you upgrade later, visits recorded while you were free are already there and appear immediately, as long as their pages were seen in the last 120 days (see "How long this data is kept" below).

What counts as one click

A click is one visit identity, not one page view and not one tap. When someone lands on your store with ?ref= in the URL, the Sprout Affiliate theme embed stores the affiliate's code plus the moment of that first click, and everything that follows is filed under that pair. So:

• Someone who clicks the same link five times over a week is one row, dated by the first click. Re-clicking deliberately keeps the original click time, because that pair is also what the cart carries.
• Someone who browses fifteen pages is still one row, with fifteen pages inside it.
• A click on a link with a different code starts a new row.

Three things have to be true for a visit to be recorded. The app embed "Sprout referral tracking" must be switched on in your theme (Online Store, Customize, App embeds).

The code in the link must belong to an affiliate who is currently active, so clicks on a link belonging to an affiliate still awaiting approval, or one you deactivated, are rejected and never appear. And the stored referral must still be inside your attribution window, which is how many days after a click a purchase still counts.

That is the Cookie duration (days) field on the affiliate's program (30 days unless you change it, up to 365, or never expiring if you tick Infinite cookie duration). Once it lapses in the visitor's browser, page reporting for that visit stops.

Reading a row

Clicking anywhere on the row except a link opens "See pages". Every row opens, whether or not it became an order.

The page journey

Expanded, a visit lists each page with the time of day it was first seen, the page title, and the path. Three limits change how you read that list:

• Repeat views of the same page are not recorded again. One row exists per page per visit. If a buyer goes product, cart, back to product, you see two entries, not three, and the product's time is the first time they saw it. The journey is the set of pages they reached, in the order they first reached them, not a full clickstream.
• The query string is part of the page. The query string is everything after the ? in a web address, so /products/hat and /products/hat?variant=2 are stored as two separate entries. Filtered collection pages, and links carrying a ?ref_source= tag, can pad the count this way.
• A visit stores at most 200 pages, and the expanded list shows the first 50. The count in "See pages (N)" is the true stored count, so a long visit can say 60 and list 50. Past 200 pages the report is accepted and nothing further is stored.

Because the same identity keeps reporting until your attribution window runs out, a journey can also include pages from a later return visit by the same browser.

Time on store, and why it is a floor

Time on store is measured from the recorded page times, with idle gaps removed. A gap of more than 30 minutes with no new page ends that session, and a later page starts a fresh one. Without that rule, a visitor who parks a tab counts every idle hour: one real visit came out at 183 minutes and tripled that store's average on its own.

• A visit with one page shows no duration. There is no "they left" event in a browser, so a single page has no measurable length.
• Every duration is an underestimate. The time spent on the last page of a session is never counted, and time spent re-reading a page already recorded adds nothing.

"Came from: not recorded"

"Came from" is the address of the page the visitor was on when they clicked, and the browser is the only thing that can report it. Sprout Affiliate stores it when the browser hands one over and that page is on another site. Your own storefront pages are blanked, both in the theme embed and again on the server, so moving around your store cannot overwrite the real source. Only the first outside page of a visit is kept.

"Not recorded" means the browser sent nothing. That happens when the visitor typed or pasted the link, opened it from an app or a message that strips the sending address, came from a site whose settings tell browsers not to pass that address on, or browsed in a private window or with tracking protection turned on. This is not recoverable. No app can work out afterwards where a click came from if the browser never said.

The fix is to label the link at the source instead. With the link builder enabled (Settings, Affiliate links, "Link builder: link to any page, and tag where it was posted", a Professional feature), affiliates fill in the Source field on the "Build a link" card in their portal, and the builder adds ?ref_source=youtube to their link.

That tag is captured with the referral, so the order's referral detail shows "Channel tag: youtube" and the affiliate's own commission table gains a Source column. Note the difference: a source tag tells you where the affiliate posted the link, and it shows on the order. It does not fill in the "Came from" line here, which only ever reflects what the browser reported.

Filters, dates, and paging

The range picker offers Today, Last 7 days, Last 30 days (the default), Last 90 days, Last 12 months, Since last payout, All time, and Custom. Two dropdowns narrow the list: one picks a single affiliate (All affiliates by default), the other picks an outcome (All outcomes, Turned into an order, or No order). Paging appears as soon as there are more rows than the page size, which you set to 10 (the default), 20, 50, or 100.

The date range is matched against the click, not the page views. A visit that started the day before your window is not listed even if the person browsed inside it, because a returning buyer keeps reporting under the original click and dating them by their latest page would move old clicks onto the wrong day. The reverse is deliberate too: a visit that ends in an order after the window closes still shows its order badge.

One gap to be aware of. If an affiliate's link code was renamed, visits recorded under the old code show that old code text instead of the affiliate's name, are not clickable, and appear only under "All affiliates". Filtering to that affiliate by name will not include them. The commission is unaffected, because an order carrying an old code is still resolved to the affiliate who owns it now.

How long this data is kept

Each page row is deleted 120 days after that page was seen, by the housekeeping job that runs every hour. Treat this page as a rolling 120-day view, not a permanent log: even the All time range stops at 120 days.

The order badge depends on a separate copy of your orders that keeps 90 days, with orders that still owe commission kept longer. That copy is trimmed by the full order sync, which runs the first time the app loads your orders more than six hours after the last one, and always before a payout. So a visit between 90 and 120 days old can still be listed while its badge has dropped back to "No order". The commission itself is unaffected, because it lives in the affiliate's ledger, not in that copy of your orders.

When the list is empty or thin

• The theme app embed is off, or you switched themes and did not re-enable it on the new one. Nothing is recorded while it is off, and it cannot be backfilled.
• The buyer used a discount code instead of a link. Code-only orders have no click, so they never appear here, though they still earn commission.
• Ad and content blockers, private windows, and cleared site data all stop the report. Expect this page to under-count real traffic rather than over-count it. The click identity lives in the visitor's own browser, so the same person on a phone and a laptop is two visits.
• Bursts of referred traffic are capped: 60 page reports a minute from one visitor's internet address, and 600 a minute per store. Reports over the cap are dropped, so a spike can lose some page rows. That address is counted in memory only and never stored.
• A page reported against a click stamped more than 130 days in the past, or more than a day in the future, is rejected as implausible.

The same page rows appear on a single order's referral detail under "Pages they visited", trimmed to the span between the click and the order, because pages browsed after the purchase are not part of how that order happened.

On privacy: a page row holds the store, the affiliate code, the click time, the time that page was seen, the page path, the page title, and the outside page that sent the visitor. No IP address, no cookie, no browser fingerprint, no customer identity. That is why this page tells you what was looked at, not who looked.