Reading your analytics without fooling yourself
Four numbers, what each one can and cannot tell you, and the three mistakes that make creators change a page that was working.
A dashboard is a machine for producing confident conclusions from thin evidence. Here is how to use yours without falling for it.
Barelink gives you four headline numbers: views, visitors, clicks, and click-through. They answer different questions, and mixing them up is the most common way to end up changing a page that was fine.
What each number is for
Views count page loads. They tell you whether your posting is working. If views move, something you did off the page caused it. They say nothing about whether the page is good.
Visitors count distinct browsers. The gap between views and visitors is repeat interest. A visitor-to-view ratio near 1:1 means people arrive once and leave; a ratio closer to 1:3 means people come back, which usually means the page is a hub they return to rather than a door they pass through once.
Clicks count taps on your links. This is the only number that corresponds to something you want.
Click-through is clicks divided by views. It is the one to watch when you change the page, because it isolates the page from your posting. Views can triple because a clip went out; if click-through holds while views triple, the page handled the extra traffic fine.
The three mistakes
Reading a day as a trend. A single day with 40 views can swing click-through by twenty points on the strength of four people. Compare seven-day blocks, not days, and compare the same weekdays, because weekend traffic behaves differently from Tuesday traffic on most accounts.
Changing two things at once. If you reorder the links and swap the video, and click-through improves, you have learned nothing about either. One change per comparison window. This is slower and it is the only version that produces knowledge.
Trusting referrers more than they deserve. Referrer data is missing or collapsed for a large share of in-app traffic. Apps strip it, and some route through their own domain. "Direct" is not a channel, it is a bucket for everything that did not announce itself. Use referrers to spot something you did not expect, never to conclude that a platform is not sending traffic.
The breakdowns that earn their place
Per-link clicks are the most actionable thing on the page. If your second link outperforms your first, that is not a curiosity. Swap them and check again next week.
Countries matter for timing more than anything else. If a third of your audience is eight hours behind you, your posting schedule is leaving money on the table regardless of what the page looks like.
Devices are usually a sanity check: if desktop is more than a small minority, your assumptions about how the page is seen are wrong, and you should look at it on a desktop yourself.
A weekly routine that takes five minutes
Same day each week. Write down four numbers: views, click-through, top link, top referrer. Note in one line what you changed since last week, if anything.
After a month you will have four rows. Those four rows will tell you more than the dashboard did on any single day, because they contain the one thing a dashboard cannot show you: what you did, and what happened next.