Tracking links, monthly reports and simple A/B tests

Likes are easy to count and hard to defend. The second month in a social media job usually brings the question "what did this actually do for us?", and the answer depends on tracking you set up in the first month.

UTM links

Analytics tools credit a visit to a source only if the link says where it came from. A UTM link adds that information as parameters:

https://www.example.com/typing-test?utm_source=instagram&utm_medium=social&utm_campaign=typing_launch&utm_content=story_day1

• utm_source: the platform (instagram, linkedin, facebook).
• utm_medium: the channel type; use "social" for organic posts and "paid_social" for ads.
• utm_campaign: the campaign name (typing_launch).
• utm_content: which post or placement, so two posts in one campaign can be compared.

Keep everything lowercase with underscores, and keep a shared sheet listing every link you create, because "Instagram" and "instagram" become two sources in reports. Shorten the link if the platform shows it. For profile links, change utm_content when you change the post so you know which one drove the visits.

The monthly report

A report is read by someone with five minutes. Put the answer first, in plain words: "Social sent 1,240 visits this month, up from 900, and 38 of them signed up." Then four short sections: what we posted (count by platform and format), what reached people (reach and engagement rate, with last month beside it), what it did for the business (clicks and sign-ups from UTM data), and what we will change next month. Add one chart at most. Worked example of a decision line: "Carousels averaged twice the saves of single images, so next month three of the five weekly posts will be carousels." A report that ends without decisions is a scoreboard.

Simple A/B tests

Test one variable at a time: the first line, the image, the posting time, or the call to action. Keep everything else the same, post both versions in the same week at similar times, and judge by the metric that matches the goal, not whichever number looks better afterwards. Example: two versions of the same post, one opening with a question and one with a statistic, both at 9 am on weekdays, measured by link clicks from the UTM data. If the difference is small, call it a draw; a handful of likes is noise. Log the result with the date so the team stops re-running the same tests.

Common mistakes and a practice habit

• Posting plain links and then being unable to show what social delivered.
• Mixing capital letters and spellings in UTM values.
• Reporting follower count as the headline number.
• Changing the image and the caption at once and calling it a test.

Practice habit. Build a UTM sheet with ten links for an imaginary launch across three platforms. Then write a one-page report for a month you invent, ending with three decisions.