You published one article and shared it everywhere: Facebook, LinkedIn, Instagram, X, maybe Threads. A week later someone asks the obvious question: which channel actually sent people to read it? If your honest answer is "I'm not sure", you're not alone. Most businesses share content widely but measure it narrowly, usually by glancing at likes and comments rather than actual visits to the article itself. Likes don't pay the bills and they don't tell you where your readers came from. This guide walks through exactly how to find out which social channel is genuinely driving article visits, what tools you need, how to set them up correctly and which numbers to trust (and which to treat with suspicion).
Key takeaways
- Engagement metrics (likes, comments, shares) are not the same as article visits; you need referral and traffic data to know what actually happened after someone clicked.
- UTM parameters are the single most reliable way to attribute a visit to a specific channel, post or campaign, and they work across almost any analytics platform.
- Platform-native analytics (Facebook Insights, X analytics, LinkedIn post stats) show clicks on the platform, but not always confirmed landing-page visits, so cross-check with your website analytics.
- Dark social (private messages, group shares, copy-pasted links) will always under-report in referral data; treat 'direct traffic' spikes after a post as a clue, not proof.
- Compare channels over several posts and weeks rather than one article, since a single post's performance is noisy and can mislead you about which channel is actually strongest.
- Set up UTM tagging and a simple spreadsheet or dashboard before you need the data, not after someone asks for a report.
Why likes and comments don't answer this question
It's tempting to judge which channel 'worked' by which post got the most reactions. But a post can get plenty of likes without anyone leaving the platform to read your article, and a quiet, unremarkable post can quietly send several genuine readers through. Engagement and traffic are different behaviours. Someone liking a LinkedIn post has told you they approve of the headline or the image. Someone clicking through to your article and spending two minutes reading it has told you something far more useful: that the content was worth leaving the platform for.
If your goal is to understand which channel drives article visits, you need to measure the click-through and the landing, not the on-platform reaction. That means shifting your attention from the social platform's engagement numbers to your website's traffic data, specifically the referral source of each visit.
UTM parameters: the simplest reliable method
A UTM parameter is a short bit of text added to the end of a URL that tells your analytics tool where a visitor came from. For example, if your article lives at yoursite.com/article, you might share it on Facebook as yoursite.com/article?utm_source=facebook&utm_medium=social&utm_campaign=article_launch, and on LinkedIn as the same URL with utm_source=linkedin instead. The article itself doesn't change. The link does.
Once someone clicks that link and lands on your site, most analytics platforms (Google Analytics 4, Plausible, Fathom, Matomo and others) will record the utm_source, utm_medium and utm_campaign values against that visit. This means you can later filter your traffic report by source and see, in black and white, how many visits came from facebook versus linkedin versus x versus a Threads post.
Three practical rules make UTM tagging useful rather than messy:
Keep source names consistent and lowercase. 'Facebook', 'facebook' and 'FB' will show up as three separate rows in your report if you're not careful, scattering your data pointlessly.
Use utm_medium consistently too, something like 'social' for all organic social shares, so you can group all social traffic together when needed and still break it down by source.
Give each specific post its own utm_campaign or utm_content if you're posting about the same article more than once, so you can tell your Tuesday post from your Thursday repost on the same channel.
Setting up UTM links without losing your mind
You don't need special software to build UTM links, though a few tools make it less fiddly. Google's own Campaign URL Builder is a free, simple form where you paste your article URL and fill in source, medium and campaign fields; it spits out the full tagged link ready to paste into your post. If you're managing a lot of channels and posts, a spreadsheet with columns for article URL, source, medium, campaign and the resulting tagged link saves you from retyping the same structure every time and keeps a record you can refer back to.
A common mistake is tagging the link once and then reusing the same tagged URL everywhere. If you do that, every channel's traffic gets attributed to whatever source you used first, which defeats the entire purpose. Each platform needs its own uniquely tagged version of the link.
Also worth knowing: some platforms strip or alter URL parameters for privacy or security reasons, and link shorteners can occasionally cause problems if they don't preserve the full query string. Test a tagged link by posting it, clicking it yourself (or asking a colleague to) and checking that the visit appears correctly in your analytics before relying on it for a real campaign.
Reading platform-native analytics correctly
Each social platform shows you its own numbers, and they're worth understanding, but they answer a slightly different question than 'how many people visited the article'.
Facebook Page Insights and the Meta Business Suite can show link clicks on a post, which is useful but counts clicks on the link within Facebook, not confirmed arrivals on your website (someone might click, change their mind while the page loads, or have ad blockers interfering).
X's native analytics show 'link clicks' per post for the account owner, a reasonably direct number, though it's reported by X itself and worth treating as an estimate rather than a perfectly reconciled figure.
LinkedIn shows click-through data on your own posts, particularly from a personal profile, which can be a genuinely useful signal for B2B content since LinkedIn audiences tend to click through with intent.
Threads, being a newer platform, has more limited native analytics than the others; according to Meta's developer documentation for the Threads platform, available insights are narrower than on Instagram or Facebook, so treat Threads numbers as directional only.
The sensible approach is to use platform-native numbers as an early, rough signal (how many people even clicked), and use your website analytics with UTM tagging as the source of truth for how many of those clicks became genuine visits.
Dark social and why your numbers will never be perfect
No matter how carefully you tag your links, some traffic will always show up in your analytics as 'direct' when it actually originated from social media. This happens when someone copies a link out of a Threads post and sends it in a text message, when a link is shared inside a private Facebook group or Messenger chat, or when someone reads a post on their phone and later types your article title into Google on their laptop instead of clicking through immediately.
This is often called dark social, and it's a genuine limitation of all referral tracking, not a mistake you've made. The practical response is not to panic about perfect attribution, but to watch for patterns: if you notice a spike in direct traffic in the hours after you post on a particular channel, that's a reasonable (though not provable) clue that the channel is contributing more than its tagged numbers suggest. Treat it as context for your judgement, not as a number to report as fact.
Comparing channels properly: look at trends, not single posts
A single article shared once on each channel tells you almost nothing reliable. One post might land at a good time, one might get a lucky algorithm boost, one might be competing with a big news story that day. None of that reflects the underlying strength of the channel for your audience.
A more useful approach is to track UTM-tagged visits across several articles and several weeks, then look at the pattern rather than any one data point. Build a simple table: channel, number of posts shared, total tagged visits, average visits per post. Over time this tells you, for example, that LinkedIn consistently sends fewer but more engaged visitors (longer time on page, lower bounce), while X sends a larger number of quick, shallow visits. Both pieces of information are useful, but only if you're looking at enough data to trust the pattern.
It's also worth separating organic posts from anything paid or boosted, since mixing the two will make an underperforming organic channel look artificially strong, or vice versa.
Putting it into a simple weekly habit
None of this requires enterprise software. A workable system for a small business or small marketing team looks like this: before sharing an article anywhere, create a uniquely tagged link for each channel using a consistent naming structure. Share the tagged links, not the plain article URL. Once a week, check your website analytics' referral or acquisition report, filter by UTM source, and note the visits for each channel against each article. Keep this in a running spreadsheet so you can spot patterns over months, not just days.
If you're using a content platform that creates and schedules both your article and its social versions, check whether it applies UTM tagging automatically when it repurposes an article link for each channel. If it doesn't, you can add the tagging yourself using the same principles above. Either way, the measurement itself depends on your website analytics tool, since that's where the actual visit is recorded, not the social platform.
Frequently asked questions
What's the difference between UTM source and UTM medium?
utm_source identifies where the traffic came from specifically, such as facebook, linkedin or x. utm_medium describes the general category of that traffic, such as social, email or referral. Using both consistently lets you look at all 'social' traffic together or drill down into one specific platform.
Can I measure article visits without using UTM parameters?
Your analytics tool will usually still record a referrer when someone clicks a link directly from a social platform's app or website, so you'll see some channel data without tagging. However, referrer data is far less reliable than UTM tagging: it can be blocked by browsers, lost when platforms redirect links through their own shorteners, or simply missing when an app doesn't pass referrer information. UTM tagging gives you a consistent, deliberate record instead of relying on what each platform happens to pass along.
Why do my social platform's own numbers not match my website analytics?
They're measuring different things. A platform's link-click count often includes clicks that didn't finish loading, clicks from bots or previews, or clicks that got interrupted. Your website analytics only counts a visit once the page actually loads in a browser with analytics tracking active. Expect the two numbers to be in the same ballpark, not identical.
Should I bother tracking Threads given its analytics are limited?
Yes, but rely on UTM-tagged website visits rather than Threads' own reporting. According to Meta's developer documentation for the Threads platform, native insights are currently narrower than on Instagram or Facebook, so your own website analytics, filtered by a threads UTM source, will usually be the more complete picture of what that channel is actually doing for you.