Clicks, unique clicks, traffic sources and bots: what your statistics really mean
The difference between a click and a unique click, why a repeated click is not counted, what “Direct” means as a traffic source and which bots are excluded.
A number without an explanation misleads more easily than no number at all. You got 1,000 clicks: is that 1,000 people? Usually not. Here is what each column in your statistics actually counts.
Click and unique click
A click is one opening of the link, counted by the application. A unique click is the first opening of the link from a given IP address: if the same address opens it again, it shows up only as a click, not as a unique one. Two useful consequences follow: several colleagues in an office who go online through the same address may show up as a single unique, and the same person, once on wifi and once on mobile data, may show up as two.
Why a repeated click is not counted
The application sets a short, 15-minute cookie in the browser each time a link is opened. If the same browser opens the same link again within that interval, the opening is not added to the statistics. A person who taps the link three times in two minutes therefore counts as a single click.
Traffic source: where the click came from
The traffic source (referrer) is the page or app the link was opened from. When it is empty, it shows up as “Direct”: the link was opened from an app that does not pass on the source (some messaging or email apps) or was typed in. That is why part of your traffic appears as “Direct” even though it came from a campaign.
Bots
Public links are opened by programs too, not just by people: search engines, the apps that show a preview when you paste a link into a message (Facebook, WhatsApp, LinkedIn, Slack, Telegram and others), automated tools. Li-Nk recognises known bots by the identifier they send and does not count them as clicks, so a preview does not show up in the statistics as if someone had tapped the link.
Recognition works from a list, so it has a limit: a program that poses as an ordinary browser, for example some security scanners, may go unnoticed and show up as a click. A small margin of noise remains with any service of this kind.
How to read the numbers
- compare trends (this week against last week), not absolute values;
- use unique clicks when you want to know roughly how many different places reached you;
- use clicks when you want to know how often the link is used.