A social post can attract attention without sending many people to the article. A link can also attract clicks without producing much useful on-site engagement. These are different steps, and each needs its own measure.
Start with the path, not a single number
Compare the post’s link clicks with the visits recorded on the destination site, then look at what those visits do. The platform and the website may count different events, use different time windows or miss different activity. A mismatch is a question to investigate, not automatically a tracking failure.
Use a consistent tagged link for each adaptation. Google describes campaign parameters such as utm_source, utm_medium and utm_campaign; utm_content can distinguish versions. For example, keep the campaign name consistent for one article and distinguish the personal X post, news post and LinkedIn post. Test the link and reporting before relying on it.
Look beyond arrival
A session is a visit measure, not a person and not evidence that an entire article was read. Google Analytics defines an engaged session using conditions that include duration, a key event or multiple page or screen views. That is a useful signal of activity, but it still does not reveal comprehension or guarantee that someone read the article.
For an editorial comparison, review landing-page sessions, engaged sessions and engagement rate alongside a clearly defined next action, such as a voluntary newsletter signup if it is actually measured. Keep the article, time window and attribution scope visible. A small audience can make percentages unstable; report counts as well as rates.
What a dashboard cannot tell you
A high engagement rate on a handful of visits is not a winner. A lower rate from a broader audience is not automatically a failure. The post’s promise, topic, timing and audience can all differ. These comparisons suggest what to investigate; they do not establish why readers behaved as they did.
Missing source information also matters. Google says a session can be processed as direct when referral information is unavailable or excluded. Some private shares may be hard to attribute, but direct traffic is not a count of private shares. Tracking settings, consent and the tools in use further limit the picture.
A useful first review is therefore modest: check the tagged links, compare post clicks with landing-page sessions over the same period, and examine on-site signals and their limitations. Then test a specific editorial change. The aim is not to manufacture a perfect reader score; it is to make better decisions with incomplete evidence.
