What link clicks measure
A link click is recorded when someone taps a part of the ad that is meant to take them somewhere: the headline, the call-to-action button, the display link or the linked image. Meta counts every such tap, including repeat taps by the same person, so the figure is clicks rather than people. The companion column, unique link clicks, counts people instead.
The destination does not have to be your website. A tap that opens a message thread, an app store listing, an Instant Experience or an event page all counts as a link click. That is why the number can look healthy on a campaign sending almost no traffic to your site — the clicks are real, they simply end somewhere else.
Why link clicks matter
Link clicks are the unit Meta charges against when you buy on a cost per click basis, so the number sits directly under your media cost. It is also the first honest sign of interest in an offer: an impression is something that happened to a person, a link click is something they chose.
For a small advertiser it is often the only metric with enough volume to compare ads within a few days. Conversions arrive slowly and unevenly; clicks arrive quickly, which makes them a fair early guide to which creative deserves more budget — provided you never mistake them for results.
Common mistakes with link clicks
The biggest is treating link clicks as website visits. They are not. A tap can be abandoned before the page appears, a weak connection can drop it, and a click into WhatsApp never touches your site at all. Comparing them against landing page views shows the size of the gap.
The second is expecting them to match your analytics. GA4 counts sessions on your own site, with its own timing and consent rules, and will almost always report a smaller number. Neither tool is broken; they are counting different events at different moments. Reconcile the trend, not the total.
The third is optimising a lead or sales campaign for link clicks because clicks are cheap. Meta will faithfully find people who tap things, and that is a different population from people who buy.
How to act on it
Use link clicks as a diagnostic, not a target. Pair them with cost per link click to judge auction pressure, and with landing page views to judge whether the site is receiving the traffic you paid for. When the two diverge sharply, treat it as a website problem first: page speed, redirects, or a page that fails on a slow mobile connection.
If you want to know how many people left Meta entirely, read outbound clicks instead, and keep the campaign objective aligned with the outcome you actually want. In Nepal, where much of the audience is on mobile data with variable speeds, the distance between a click and a loaded page is wider than most dashboards suggest. It is worth measuring rather than assuming.