BlogMeasurement

Which Social Media Metrics Matter for a Small Business

By the Postd team ·

Track the step you are trying to fix, not the number that is easiest to count. A customer passes four gates before paying you, each one leaks, and one number per gate tells you which leaks worst. Four numbers, checked monthly, beat forty checked daily.

A leaking funnel drawn in Postd's style: four gates narrowing down the canvas, with a coral bar marking the third gate where the sharpest drop happens.

Why the dashboard does not help

Every platform hands you forty numbers and no opinion about which one matters. Faced with that, most owners look at the largest number, which is usually reach, and take from it a vague feeling that things are going well or badly.

The problem is not the quantity of numbers. It is that they answer questions about different stages, and you only have a problem at one of them. Tracking reach while your actual problem is that nobody books is like weighing the ingredients when the oven is broken.

So the useful question is never which metrics to track. It is which step of your own funnel is currently failing, and which single number describes it.

The four steps a customer takes

Between never having heard of you and paying you, someone passes four gates:

  • Seen. The post appeared in front of a human being.
  • Stopped. They read it instead of scrolling past.
  • Interested. They looked at your profile, your site, or your listing.
  • Acted. They called, booked, messaged, or bought.

Each gate leaks. Your job is to find which one leaks worst, and that takes one number per gate rather than forty across all four.

One metric per step, and only one

Seen: impressions or reach. Useful only as a denominator for the next number, never on its own.

Stopped: engagement rate, meaning interactions divided by reach rather than raw likes. Raw likes rise simply because reach rose, which tells you nothing about the post.

Interested: profile visits and link clicks. This is the first number that represents genuine intent, and it is the one most small businesses never look at.

Acted: enquiries, bookings, or sales. Count them by hand if you have to. A tally on paper beats an untrustworthy dashboard.

Which numbers are genuinely vanity

Follower count is the famous one, and the criticism is fair for a local business. A thousand followers in the wrong city is worth less than eighty in the right postcode.

The subtler traps are totals with no denominator. Total likes this month rises when you post more often, which measures your effort rather than your effect. Any number that improves purely because you did more of the same thing is describing your input.

The clearest example is a post that reaches thirty thousand people because a platform pushed it to strangers. It looks like the best week you have had. If none of those people live near you or want what you sell, the engagement rate collapses and the enquiry count stays flat, which is the truthful reading of what happened.

What a useful baseline looks like

A single month tells you almost nothing, because a good week and a slow week differ more than most changes you will make. Three months of the same four numbers gives you a range, and a range is what makes a movement meaningful.

Write them somewhere permanent, one row a month, four columns. That habit is worth more than any analytics feature, because it survives platform changes and it is yours.

How often to actually look

Monthly, for twenty minutes. Weekly checking encourages reacting to noise, and daily checking encourages posting for the graph rather than for a customer.

The exception is a launch or a promotion, where the point is a short burst and you want to know within days whether it landed. Outside that, the monthly rhythm suits the monthly planning pass you are already doing.

What to change when a number moves

Match the fix to the gate. Weak reach is a frequency, timing, and format problem. Weak engagement rate against decent reach is a subject and opening-line problem. Good engagement with few profile visits means the post entertained without ever suggesting a next step.

Change one thing at a time and give it a month. Changing three at once means the next movement tells you nothing about which one caused it.

Worked through: a bakery sees steady reach, an engagement rate that halved over two months, and unchanged enquiries. Reach rules out a distribution problem, so the leak is at the second gate. Looking back, every post that quarter opened the same way and showed the same counter. The fix is the subject, not the schedule.

The honest limits of attribution

No small business tool can draw a clean line from a post to a sale. People see a post, remember you weeks later, search your name, and arrive through a link that credits search. Any dashboard claiming otherwise is guessing confidently.

Two cheap habits close most of the gap. Ask every new customer where they found you and write the answer down. Watch whether direct and branded search traffic climbs while you post consistently, because that is usually where social shows up.

How Postd reports it

Postd keeps the reporting deliberately small: what went out, on which accounts, and how each post performed, in one place rather than four dashboards.

The aim is the twenty-minute monthly review described above rather than a wall of charts. Posts are prepared for you and published after you approve them, and nothing publishes without your approval.

What it deliberately avoids is claiming a clean line from a post to a sale, for the reason given above. The features page covers what is reported.

Your next step: write down this month's four numbers. In three months you will have a baseline, and the arguing stops.