Measuring Social Media Without Clicks

The monthly social report lands with a few sessions, no conversions worth naming, and one form fill that probably came from somewhere else. The owner does the math on the hours and asks why the channel is still on the list. It is a fair question, and it comes up often enough that I stopped hearing it as a social media question...it is a measurement question.

The buying conversation on social does not happen on your website. It happens in the feed, in the comments under a post, and in a screenshot someone sends to a spouse or forwards to the person who signs the check.

Days or weeks later that person opens a browser and types your business name. Not one step in that sequence produces a click...the report has nothing to record.

Structural Gaps in Click Tracking

Platforms are built to keep people inside the app, and three ordinary mechanics carry that out:

• Native Reach: Posts that hold attention inside the app tend to get treated better than posts that send people away.
• App Browsers: The built in browser inside a platform behaves inconsistently from one app to the next.
• Copied Links: Much of the sharing happens by pasting a URL into a text thread, a group chat or an email.

GA4 records what the browser hands it...a session arriving with no referrer and no campaign parameters lands in direct or unassigned, and switching analytics tools does not change that arithmetic. But the gap is not a setting you forgot to switch on...it is the shape of how people pass things along.

Once you accept that, the job changes...you assemble evidence from several imperfect sources that fail in different ways, and each one is wrong in its own direction...that is exactly what makes them worth reading together.

Branded Search and Direct Visits

When social is working, it usually shows up somewhere other than the social report...it shows up as people looking you up by name. Branded queries in Search Console, direct sessions, and calls or direction requests from your Google Business Profile all move when awareness moves.

The method matters more than the tool, so pick a consistent window, something like the campaign week plus the two weeks after it, and a consistent baseline period before it. Set the window before the campaign starts, not after. Then look at the same three things inside every window you set:

• Branded Queries: Impressions and clicks on your own name, including the misspellings and near misses people type.
• Direct Sessions: Unassigned arrivals on the home page and on the main service pages.
• Inbound Actions: Calls, form starts and profile activity falling inside the same window.

But one campaign proves nothing...seasonality, a news mention (or a competitor going quiet) will all move these numbers. What you want is a pattern that repeats across several pushes. If branded search lifts in the days after you post and settles when you go quiet, that is real signal...no report will ever draw the line for you.

Better Questions at First Contact

The cheapest attribution instrument most businesses own is the moment a person first talks to them. It is also the one most often wasted...a dropdown collects the answer, and everybody picks Google because it sits at the top.

So rebuild that moment out of three cheap parts:

• Open Field: Swap the dropdown for a text box and let people answer in their own words.
• Spoken Ask: Whoever answers the phone asks out loud and writes down what was said.
• Second Question: After how did you hear about us, ask what they saw and/or read.

That second question is where you get answers like "my sister sent me your video" or "I have been seeing your posts for months." Neither answer exists anywhere in your analytics...neither one will ever be recovered by a better report.

Self reported attribution is noisy, because people misremember, they compress a long sequence into whichever touch came last, and they say Google when they mean they searched your name after seeing you somewhere else. Read it as a tally...not a ledger, and review it quarterly next to the branded search picture.

Campaign Tags for Shared Links

Some people do click, and those clicks should be legible. Campaign tagging is unglamorous work...it is also the difference between a report that says something and a report that says direct.

• Naming Convention: Lowercase, no spaces, and identical source and medium values across every channel you post on.
• Campaign Names: One name per push, used on every platform, so later comparisons line up.
• Internal Links: Never tag them on your own site, which restarts sessions and misassigns the credit.
• Shared URLs: Keep the parameters on links you expect people to pass along, since a tagged address pasted into a text message carries them intact and recovers some of what would otherwise stay invisible.
• Distinct Destinations: Where the offer allows it, point a campaign at its own landing page or a code you use nowhere else.

Discipline here is worth more than sophistication...a boring scheme applied every time beats a clever one applied whenever somebody remembers. None of it recovers the conversations that never left the app...it does keep the visible half of the picture honest.

Does the Channel Earn Hours?

Before you can judge a channel you have to say what job it has. Social rarely closes business for a local or regional company...at the end of the day it does other work, and that work is worth writing down.

• Reassurance: It steadies someone already considering you who wants another look first.
• Referral Presence: You stay visible to the people who send work your way.
• Visible Proof: Buyers who cannot judge your output from a service page can see it.
• Hiring Reach: The people who might come to work for you meet the place early.

Write the job down, then judge the channel against it. The cost is hours...track them honestly, including the ones spent on approvals and rewrites. Then read the evidence together, since branded search movement, intake mentions and the substance of the comments all count, and buying questions count for more than compliments.

If it is still ambiguous, run the quiet test. Reduce posting to a minimum for a defined period, hold everything else steady, and watch intake mix and branded search. Then start again. It is not a controlled experiment and you should not present it as one...a channel nobody misses is telling you something.

Your First Month of Work

Start with intake, because it costs nothing and it pays this month. Replace the dropdown with an open field, add the follow up question, and tell whoever answers the phone what to record. Then set your measurement window and your baseline...branded query impressions, direct sessions and intake mentions belong on one page you look at monthly.

Fix the tagging convention on the way through, and give the channel a written job. Above all, give yourself a quarter before drawing conclusions. None of it requires a new tool.

You will end up with a defensible judgement rather than a precise number, and I think that is the better trade. For an owner deciding where the next block of hours should go, the precise number was never the thing missing...the judgement was.