Which Instagram numbers matter for a small business?

It sits in the biggest font on your profile and it is the last thing to move, which is an unkind combination for anyone checking it on a Tuesday night. It counts people who agreed to something once, and some of them have not opened the app this year.

A follow happens after somebody saw a post, liked what they saw enough to tap your name, then read your profile and decided. Three things had to go right and the follower count reports only the last one. If you want it to climb, the work sits in the numbers before it.

Look at it monthly. Never per post.

Reach counts the separate accounts that saw a post at least once. Views counts every time it was shown, including the same person coming back for another look, which is usually why the two numbers do not match.

The useful part is the split, not the total. Instagram's own creator guidance says you can break Accounts Reached down by followers and non-followers "so you can see how much of your reach is coming from people who don't already follow you", and track that over time. Set a long window rather than last week, because one quiet Wednesday moves a week and moves nothing over three months.

If nearly all of your reach is people who already follow you, what you have is a newsletter. There is nothing wrong with a newsletter, and your existing followers are usually the ones who buy things. But nobody new is arriving, and a business needs the other half of that bar to exist. If the non-follower half used to be there and has fallen away, why has my Instagram reach dropped goes through the usual causes.

A profile visit is somebody tapping your name and landing on your page. It is the moment a post has finished its job and your bio takes over, and it is the number the owners I work with check least.

Read follows against profile visits, not against reach. Out of everybody who came for a look, how many stayed. That one comparison separates two problems that feel identical from the outside. Plenty of visits and almost no follows means the content is working and the profile is losing them: the fix is your bio, your highlights and the top row of your grid. Barely any visits means nobody is getting far enough to be lost, and the fix is the content.

Link taps sit one step further along, the last thing that happens before anybody can book, buy or enquire. If that number is flat, check how often you ask anybody to tap it before you blame the link.

Instagram publishes what it tries hardest to predict, and the lists differ across the app. In Reels it names how likely you are to "reshare a reel, watch a reel all the way through, like it, and go to the audio page". In Explore it says the most important actions it predicts are "likes, saves, and shares".

Nobody outside the platform knows how any of that is weighted, and anyone who tells you what a save is worth next to a like is guessing. What the lists are good for is direction. A like costs a thumb. A comment costs a sentence. A save is somebody deciding to keep your post. A share is somebody putting their own name to it in front of their friends.

So the useful move is to build the reason to keep or pass it on into the thing itself: a checklist, the answer to the question your customers ask you every week.

Watch time is how long people stayed with a video, added up. Retention is the shape of the same thing, and the shape is the half worth having. It starts with everybody who pressed play and falls away as they go, so the steepest part of that line is the second you lost them.

No other number tells you where to make the cut. A drop right at the front, before you have finished a sentence, is your opening, and why do people scroll past my reels in the first second is about that on its own. A line that holds and then sags means the opening worked and the middle is where the work is.

Every article that gives you a target either invented it or averaged a pile of accounts nothing like yours. A wedding photographer, a florist and a physio have different base rates for every number on this page, and so do two florists in different suburbs.

Published averages do exist, and there is one on my audit page with both studies it came from linked, because knowing what ordinary looks like is steadying when you are convinced you are the only one flatlining. That is what it is for. It is not a target.

Your comparison is your own account over a longer window. One post tells you close to nothing. Eight weeks of posts tells you which pillars pull, which openings earn saves, and what happens when you skip a fortnight.

Pick a day and keep it. Open Insights, set the window wider than the last seven days, and look at four things: the follower and non-follower split of your reach, your profile visits, your follows against those visits, and the retention curve on the video that did least.

Then write two lines somewhere you will find them again. What moved, and what you did differently. Do that for a couple of months and you will have the only report about your account that is about your account. Checking it hourly does nothing except make you miserable, and I have done that too. If you want the rest of the routine written out, the pillars, the batching session and the weekly checklist, it is all in my free guide Scroll Less, Grow More.

Every one of these is a symptom. They tell you what happened and never why, and the why is a judgement somebody makes with the posts open beside the numbers.

That is what an audit is for. You send your handle, screenshots of your Insights and a few lines about the business, about ten minutes of work, and I sit with the numbers and the posts together. What comes back is which of your numbers is saying something, the post or the pattern behind each one, and what to change first, in order, with nothing in the document I cannot point at a post or a figure to justify. It is A$300 once, no membership attached, and it lands within five business days of your details reaching me, written as your own private page that stays yours. The Doomscrollers Handbook, A$50 on its own, comes with it. There is a full example report on the site if you would rather read one before you decide.